STUDIES IN ANCIENT GREEK
SOCIETY
STUDIES IN
ANCIENT GREEK
SOCIETY
«*•
THE PREHISTORIC jEGEAN
GEORGE THOMSON
LAWRENCE & WISHART
LONDON
1949 '
COPYRIGHT
Printed in Great Britain by
ame ’° l Press Ltd., London and Southampton (t.u.)
To the memory of
HUGH FRASER STEWART
TToO eTv’ f) ’AAi^Osioc; pf)v ttAovov gas
PaSiovoTyra Aoyia Taya'
Ttjv rniyr) ttis 6£v tt] PpluKEts
Psctoc ctou, J, Av0pcoTTE, povaycx.
©oc rn Pprjs ttocvtou ctto Todpiaapcx
— ob dppa^dbvas Tan-peon^ —
tt]S KapSias ctou kcci *rou vou ctou
p£ Td TrdVTCC Tffc scoffs.
— Palamas
PREFACE
This volume is planned as the first of several with the aim of
consolidating the ground covered in JEschylus and Athens. It is in
effect an expansion of the first five chapters of that work, rein-
forcing the argument and treating at length some fundamental
problems that were only touched on. there, especially matri-
archy, land-tenure, Aigean prehistory, and epic. Its range
.coincides approximately with the Bronze Age, except that the
(evolution .of epic is followed down to its culmination in the
isixth century B.c. In the second volume I hope to deal with
the growth of slavery a nd th ejarigins . of. science.
The task I have set myself is to reinterpret the legacy of
Greece in the li ght of Marxism. Some of my critics seem to
think that treated in this way Greek studies lose their value.
I believe that only in this way can they recover it. Everybody
knows that for many years past their popularity has been de-
clining, and the reason is that they have lost touch with the
forces of human progress. Instead of being a message of hope
for the future, as they were in the great days of humanism, they
have become a pastime for a leisured minority striving in-
effectually to find a refuge from it. Our Hellenic heritage
must be rescued from the Mandarins, or else it will perish,
destroyed by its devotees. -
It need hardly be said that my treatment of the subject is
severely restricted by the limitations inherent in any single-
handed attempt to cover so vast a field. Recent developments
in archaeology and linguistics have made it clearer than ever
that Greek history must be studied as an episode in the general
history of the Near East, and this can only be done effectively
by collective research based on an agreed scientific method. If
my work draws attention to this need, its very shortcomings
will have served a useful purpose.
If I can do more than that — if I can convince at least my
younger colleagues that in the age in which we live the new
humanism, inherited from the old but enriched by the four
8 STUDIES IN ANCIENT GREEK SOCIETY
most eventful centuries in history, is Marxism, then it will be
for them to renew the vitality of Hellenism so that it may
exercise an influence on the future of British culture worthy
of More, Bacon and Milton.
I have been mindful, too — indeed it has been impossible to
forget — that, while I have been writing this book, the Greeks
have been fighting for liberty with a heroism unequalled even
in their history. That is why I have inscribed those lines from
Palamas. They express one of the profound truths of Marxism
in the words of a poet who more than any other spoke for
the people of modern Greece, voicing their determination to
be free from all forms of oppression, free too from, the domina-
tion of the past, while proving their fidelity to it by their
creative energy in building a new Hellas.
I wish to thank the following for permission to reproduce
illustrations of which they hold the copyright: Messrs. Mac-
millan and Co. (figs. 18, 20, 23, 25, 26, 30, 42, 54, 57, 60, 68,
70, 71 , 73); the Syndics of the Cambridge University Press (figs.
3 » 5 ) 8, 9, 33, 36, 39, 40, 43); and the Editors of The Classical
Review (fig. 44). For the remainder of the illustrations I am
indebted to Mrs. H. F. Stewart.
I must also express my gratitude to Dr. N. Bachtin for the
unfailing stimulus of innumerable discussions ranging over the
whole subject-matter of the book long before it assumed book
form; and for assistance and advice on various problems to
Dr. F. J. Tritsch, Mr. Rodney Hilton, Mr. Alick West, Mr.
John Irwin, Mr. R. F. Willetts, and Mr. O. M. Thomson.
June 1948. George Thomson,
CONTENTS
PAGE
Introduction 21
PART ONE
KINSHIP
Chapter I. Totemism
' 1, The Comparative Study of Ethnology and Arch-
aeology . . . . • .33
2. The Origin of Totemism . . . .36
3. The Origin of Exogamy . .... 41
4. The Totemic Cycle of Birth and Death . . 45
5. From Totemism to Religion .... 49
6. Totemism. in Palaeolithic Europe ... 52
Chapter II. The Nomenclature of Kinship
1. Structure of the Tribe ..... 5.8
2. The Classificatory System: Type I •. .60
3. Ritual Promiscuity . . . . .66
4. The Classificatory System: Type II . . .67
5. Group-marriage ...... 69
6 . Decay of the Classificatory System . . .71
7. The Descriptive System .... 78
Chapter III. From Tribe to State _
1. The League of the Iroquois . . ' . - . 87
2. The Roman Tribal System .... 92
3. Matrilineal Succession of the -Roman Monarchy . 97
4. The Populus Romanus ..... 99
10
STUDIES IN ANCIENT GREEK SOCIETY
Chapter TV. Greek Tribal Institutions page
1. ^Eolians, Dorians, and Ionians . . . .102
2. The Attic Tribal System . . . .104
3. The Household . . . . . .109
4. Pre-Hellenic Clans in Attica . . . .112
5. Totemic Survivals: Snake-worship . . .114
6. Totemic Survivals Clan Emblems . . .120
7. Clan Cults and State Cults . . . .123
8. The Clan Basis of the Eleusinian Mysteries . 127
9. The Treatment of Homicide . . . . 132
10. The Law of the Heiress . . . . .137
11. Ancient Greek Ethnology .... 140
12. Linguistic Evidence of Matrilineal Descent . .144
PART TWO
MATRIARCHY
Chapter V. The Matriarchal Peoples of the JEgean
1. What is Matriarchy? ..... 149
2. The Lycians . . . . . . 163
3. The Carians and Leleges . . . .166
4. The Pelasgoi ...... 171
5. The Minoans ...... 177
6. TheHittites . . . . . .179
7. The Legend of the Amazons . . . .180
8. The Minyai . . . . . .183
9. Some Matriarchal Survivals . . . .199
Chapter VI. The Making of A Goddess
1. Childbirth and Menstruation . . ... 204
2. Moon-worship . . . . . .210
3. The Moon in Popular Greek Religion . .214
4. Herbal Magic ...... 218
5. The Thesmophoria and Arrhephoria . . . 220
6. Rites of Ablution ...... 223
7. The Daughters of Proitos .... 226
8. Greek Goddesses and the Moon . ... 228
9. The Rape of Persephone . . . .231
10. The Female Figurine ..... 237
CONTENTS
II
Chapter VII. Some Matriarchal Deities oj the JEgeati page
1. Demeter ....... 249
2. Athena ....... 257
3. The Ephesian Artemis ..... 269
4. The Brauronian Artemis .... 276
5. Hera ........ 280
6. Apollo ....... 293
PART THREE
COMMUNISM
Chapter VIII. The Land
1. Beginnings of Private Property
2. The Problem of Ownership in Early Greece
3. Primitive Land-tenure ....
4. The English Village Community
5. Greek Husbandry .....
6 . Modem Greek Land-tenure
7. The Open-field System in Ancient Greece .
8. Redistribution of the Land
9. The Method of Distribution .
10. The Growth of Privilege
Chapter IX. Man’s Lot in Life
1. Occupational Clans ....
2. The Moirai as Spinners ....
3. The Horai and Charites
4. The Erinyes .....
5. The Indo-European Origin of the Moirai .
6. The Transformation of Moira .
Chapter X. The Formation of Towns
1. Thucydides on Primitive Greece
2. Formation of Towns in the Historical Period
3. From Tribal Camp to City-State
4. Phaeacia and Pylos ....
5. Early Athens .....
297
299
302
3°5
308
31 1
313
323
332
334
339
341
342
345
348
349
351
359
362
12
STUDIES IN ANCIENT GREEK SOCIETY
PART FOUR
THE HEROIC AGE
Chapter XI. The Mycenean Dynasties page
1. The Traditional Chronology .... 3.69
2. The Archaeological Framework . . ' . .371
3. The Traditional Dynasties .... 374
Chapter XII. The Achceans
1. Distribution of the Achaeans . . . 385
2. The Aiakidai . . . . . .387
3. The Ionians ...... 390
4. The Peloponnesian Achaeans . . . .392
5. The Origin of the Achaeans . . . *395
6. The Pelopidai ...... 400
Chapter XIII. The Clash of Cultures
1. The Social Character of the Achaeans . . 412
2. The Homeric Treatment of the Matriarchate . 416
3. The Kingdom of Odysseus . . . 420
4. The Leleges of Western Greece . . . 425
5. The Superiority of the Achaeans . . . 430
PART FIVE
HOMER
Chapter XIV. The Art of Poetry
1. Speech and Magic . . . . .435
2. Rhythm and Labour ..... ,445
3. Improvisation and Inspiration .... 454
Chapter XV. The Ritual Origins of Greek Epic
1. The Problem ...... 463
2. The Strophe ...... 464
3. The Hexameter ...... 474
4. The Chorus . . . . . . 479
5. The Epic Prelude ...... 490
6. Songs after Supper 494
CONTENTS
13
Chapter XVI. Homeric Archeology and Linguistics
PAGE
I. Datable Elements .
• • •
5°i
2 . The Mode of Burial
503
3 . Helen ....
505
4 . The Epic Dialect .
515
5 . The Epic Style
527
Chapter XVII. The Ilotneridai
1 . Aiolis and Ionia
541
2 . Homer's Birthplace
547
3 . From Court to Market-place .
549
4 . The Homeric Corpus
552
5 . The Cyclic Poems
559
6 . Diffusion of the Iliad and Odyssey
564
7 . The Recension of Pcisistratos .
571
8 . The End of Epic .
575
9 . Structure of die Iliad and Odyssey
577
Bibliography
583
Index to Maps .
601
General Index .
609
ILLUSTRATIONS
HG.
PAGB
I. Chamois dance: palaeolithic stag’s horn
Burkitt P 208
•
55
2. Athena and snake: relief from Melos .
Roscher LGRM 1. 690.
•
1 16
3. Burial mound and snake: Attic vase .
Harrison P 328.
•
1 17
4. Feast of the dead: Laconian relief
Roscher LGRM 1. 2567.
•
118
5. Ox-head on shield: Attic vase .
CAH Plates 1. 283
•
121
6. A Philistine: Egyptian painting
ABS 8. 185
•
164
7. Amazon: Attic vase ....
Furtwangler pi. 166.
•
181
8. Cult of the moon: Minoan gem
Harrison T 190.
*
211
9. A Maenad: Attic vase ....
Harrison P 399.
•
217
10. Goddess with pomegranate: Attic statue
CAH Plates 1. 208.
•
219
11. Woman sacrificing a pig: Attic vase .
Harrison P 126.
•
221
12. Girls at a well: Attic vase
Baumcistcr 1. 357.
•
223
13. Artemis and Aktaion: Attic vase
CAH Plates 2. 33.
•
225
14. Three-faced Hekate: gem
Roscher LGRM 1. 1909.
•
230
15. Persephone in Hades: Attic vase
Roscher LGRM 2. 1343.
•
233
16. Venus of Willendorf: palaeolithic figurine .
CAH Plates 1. 8.
•
237
17. Thessalian figurine: terracotta from Sesklo .
CAH Plates 1. 112.
•
238
ILLUSTRATIONS
15
FIG. PAGE
18. Minoan figurine: terracotta from Knossos. . . 239
Evans PM 1. 52.
.19. Cycladic figurine: marble .... 243
CAH Plates 1. 115.
20. The ‘gesture of benediction’: bronze from Knossos 244
Evans PM 1. 507.
21. Trojan face urn ...... 250
Cook Z 3. 192.
22. Child suckled by goat: Minoan seal . . .251
ABS 9. 88.
23. Minoan double axe: intaglio from Knossos . .251
Evans PM 2. 619.
24. Minoan bull fight: intaglio from Knossos . .252
Evans PM 1. 377.
25. Mycenean cult scene: gold ring from Myceme . 252'
Evans PM 2. 340.
26. Dance at a sacred tree: gold ring from Mycenae . 253
Evans PM 1. 161.
27. Ascension of Demeter: terracotta relief . . 253
Roscher LGRM 2. 1359-60.
28. Pappas type of Demeter: terracotta from Eleusis . 254
Famell CGS 3. 215.
29. Minoan snake priestess: statuette . . . 254
CAH Plates 1. 119.
30. Britomartis: intaglio from Lyttos . . . 255
Evans PM 2. 844.
31. Descent of the god: Minoan signet . ^ . 255
Evans PM I. 160.
32. Minoan priest: relief from Knossos . . .256
CAH Plates 1. 157.
33. Demeter and Triptolemos: Attic cup . . 256
Fiirtwangler pi. 65.
34. Athena; Attic vase . . . . .258
Baumeister 3. 1152.
35. Athena, Erichthonios and Kekrops: Attic relief . 262
Roscher LGRM 2. 1019-20.
36. Athena and the daughters of Kekrops: Attic vase . 263
Harrison P 133.
l6 STUDIES IN ANCIENT GREEK SOCIETY
FIG. ' . PAGE
37. Poseidon and Hephaistos at the birth of Athena:
Attic vase .268,
Roscher LGRM 1. 2061—2.
38. Ephesian Artemis: statuette from Ephesos . ■ . 269
CAH Plates 1. 351.
39. Mother-goddess and twins: Attic vase . . 270
Harrison P 268.
40- Artemis Orthia: ivory from Sparta . . . 272
Harrison T 1 14.
41. Zeus and Hera: Attic vase . . . . 283
Farnell CGS I. 208.
42. The Minotaur: coin from Knossos . . 2,85
Evans PM l. 358.
43. Etruscan armour: stele from Vetulonia . . 290
CAH Plates I. 327.
44. Juno and Hercules: Etruscan bronze . . . 291
CR 20. 274.
45. Apollo and Artemis: vase from Melos . . 293
Farnell CGS 4. 328-9.
46. Ploughing: Attic vase . . . . • . 308
Baumeister l. Tafel x. 13.
47. Ploughing: vase from Bari . . ’ . •. 3°9
Cook Z 3. 606-7.
48. Olive harvest: Attic vase . •. . • . 3 ID
Baumeister 2. 1047.
49. Country dance: Ionian vase . ... ■> 3 11
Lambrino pi. iv. '
50. Casting lots: Attic vase . . . . . 328
Baumeister 1, 684.
51. Charites: Attic relief . . . . 34 °
Baumeister 1. 375. . .
52. Mattes Deae: relief from Avigliano . . • ' 344
Roscher LGRM 2. 2471.
53. Venus of Laussel: palaeolithic carving . •' 344
CAH Plates 1. 9.
54. Minoan ship: seal . ■ . . . . 359
Evans PM 2. 239.
55. Gold death mask of the Shaft . Grave Dynasty. . . 37 1
CAH Plates 1. 165.
ILLUSTRATIONS
17
HG.
56. Mycenean boar-hunt: fresco from Tiryns
Hall AA fig. 74.
PAGE
372
57. Embarkation scene: signet from Tityns
Evans, PM 2. 245.
372
58. The Lion Gate of Mycenae
CAH Plates 1. 161.
373
59. Perseus and the Gorgon: Attic vase .
JHS 32 pi. vi.
U)
00
0
60. Minoans in Egypt: Egyptian painting
Evans PM 2. 740.
381
61. Aegean ship: Egyptian painting
Baumeister 3. 1595.
381
62. Ariadne, Theseus and the Minotaur: gold orna-
ment ••••••
Roscher LGRM 2. 3007
•
383
63. Mycenean lady: fresco from Tiryns .
CAH Plates 1.159.
9
418
64. Dancers: Attic vase ....
ABS 30. PI. xviii.
•
441
65. Baking to music: Boeotian terracotta .
Ehrenberg pL xib.
«
446
66. Muses: Attic vase .
Furtwangler pi. 169.'
•
462
67. Alkaios and Sappho: Attic vase
Baumeister 3. 1543.
•
472
68. Mycenean dancer: gem from Vapheio.
Evans PM 3. 69.
9
481
69. Apollo and lyre: Attic vase
Baumeister 1. 96.
m
484
70. Ladies of Knossos: fresco (restored) .
Evans PM 3. 55.
487
71. Descent of the goddess: Minoan signet
m
487
Evans PM 3. 68.
488
72. Minoan chorus: terracotta
Evans PM 3. 73.
•
73. Minoan lyre-player: Hagia Triada Sarcophagus
Evans PM 1. 440.
•
489
74. Mixed chorus: Attic vase
Baumeister 3. 1948.
•
490
B
l8 STUDIES IN ANCIENT GREEK SOCIETY
HG. PAGE
75. A rhapsode: Attic vase . . . • • 49 1
Baumeister 3. 1540.
76. Drinking party: Attic cup . . . *495
Fiimvangler pi. 73.
77. Dancing girl: Attic cup . 497
Furtwangler pi. 73.
78. Gold cup from the Fourth Shaft Grave . . 5 ° 2
Nilsson HM 138.
79. Boar’s-tusk helmet: Mycenean ivory . . .502
Nilsson HM 76.
80. Aphrodite and swan: Attic cup . . . $06
Baumeister 2. 856-7.
81. Dove-headed Aphrodite: Cyprian terracotta . . 5 I 3
Roscher LGRM x. 407.
82. Footrace: Attic vase . . . . . 53 ^
Baumeister 3. 2109.
83. Iris: Attic vase ...... 538
Roscher LGRM 2. 344.
84. Sub-Mycenean soldiers: vase from Tiryns . . 543
Nilsson HM 158.
85. King Arkesilas: Laconian cup .... 562
CAH Plates 1, 379.
TABLES
I. Terminologies of Kinship .... 62
II. Dislocations in the Amerindian Kinship
Systems 77
HI. The Indo-European Nomenclature of Kinship . 81
IV. The League of the Iroquois ... 88
V. The Sabine and Etruscan Kings of Rome . 98
VI. Prehistoric Greek Chronology . . .103
VII. Evolution of Patrilineal Succession . . 155
VIII. The Descendants of Sisyphos . . .165
IX. The Kings of Orchomenos . . . .188
X. The Lapithai 189
XI. Minyas and Tyro 194
TABLES
19
PAGE
XII. The Chronology of Eratosthenes . . . 370
Xm. The Argive Pedigree ..... 378
XIV. The Aiakidai 388
XV. The Pelopidai . . ... . 408
XVI. Perieres and Thestios .... 428
XVII. The Greek Dialects .... 523
XVm. The Homeric Corpus . . . .553
MAPS
I. The Eastern Mediterranean ... 27
II. Cults of Demeter . . . . .130
III. Prehistoric Peoples of the JEgean . .168
IV. The Dimini Culture ^Thessalian IQ . .184
V. Cults of Artemis ..... 274
VI. The Argive Plain . . . . .281
Vn. Achaean Settlements in the Peloponnese . .394
Vm. The Thessalian Achaia . . . *396
IX. The Kingdom of Odysseus . . . .421
X. Anatolia and the Black Sea . . . .511
XI. The Greek Dialects ..... 524
Xn. Greece and the JE gean . fold-in map facing 622
AID TO THE READER
The references in the footnotes are to the works listed in the
Bibliography and to Greek and Latin texts. The former are
cited by the author's name followed where necessary by the
initials of the title in roman capitals. Greek and Latin authors
and titles are cited, with a few minor modifications, in the
abbreviated forms employed in the Greek-Englisb Lexicon of
Liddell and Scott (new edition) and the Latin-English Dictionary
of Lewis and Short. References to Aeschylus are by Wecklein’s
numeration.
ADDENDUM
Page 321, note 86. Add Th. 3.88.2, Paus. 10.11.4.
These passages show that the main island was the
only one inhabited permanently throughout the
year.
INTRODUCTION
The development of a neolithic economy was rendered possible
- by the series of climatic changes that followed the close of the
last Ice Age. It began somewhere in the Middle East. As the ice
retreated in the north, the climate of this region, previously
temperate, became subtropical. The open grassland, which
had stretched almost without a break from Morocco to Iran,
was split up into semi-desert tracts intersected by strings of
green oases and river beds overgrown with jungle. The roaming
bands of hunters and food-gatherers lost their freedom of
movement. They were forced to concentrate in the more
fertile areas, together with the animals and plants on which
they lived. Thus restricted, they found that the supply of
game and fruits was limited. The old technique of hunting
and food-gathering was no longer adequate. Some means had
to be found of preserving the animals and plants by bringing
i their propagation under human control. Among the species
indigenous ttf this region were the sheep, goat and pig, all
.easily domesticated, and the wild ancestors of our wheat and
barley. The animals were herded and penned, the plants sown
.'artificially; and both were tended by human labour. Hunting
land food-gathering were superseded by stock-breeding and
tillage. Besides ensuring a regular supply of milk, meat and
grain, the new economy gave rise to a number of secondary
^techniques, such as. weaving and pottery, which resulted in
further ..improvement of living standards. The people multi-
plied. The makeshift, straggling tribal camp was transformed
into a thriving village, compact and self-sufficient, though it
was continually planting out its surplus population in new
villages founded on the same model. In this way the neolithic
economy was propagated over the whole region and beyond,
wherever cultivable soil was to be found. As the limits of
< expansion were approached, the growing pressure of popula-
| tion promoted more intensive methods of cultivation, and
i meanwhile village self-sufficiency was undermined by the
j development of exchange.
21 ' STUDIES IN ANCIENT GREEK SOCIETY
The river swamps of the Nile, Euphrates .and Tigris,
teeming with wild life of all kinds, had always attracted
hunters and fishers, but to the early cultivators they presented
formidable obstacles. The soil could only be adapted to tillage
by large-scale irrigation, which required organised mass labour
working according to a plan. Such conditions could not be
satisfied until the neolithic economy was well advanced in the
adjacent areas. On the other hand, the potential fertility of
these alluvial soils was immense. Once the obstacles had been
overcome, the way was open for an increase in population and
a rise in living standards far beyond the possibilities of the
old neolithic economy. The village was superseded by the
town. The town was not merely larger, more populous, more
luxurious. It differed in its economic basis. Its surplus of
grain and livestock was so ample that it could be bartered
regularly and extensively for timber, stone and metals from
the surrounding hill tribes, whose own village economy was
modified accordingly, becoming dependent on the town.
Economic self-sufficiency, except in outlying areas, was a
thing of the past. As trade expanded, with craftsmen, mer-
chants, middlemen of all sorts, pushing their way up and down
the valleys and across the intervening deserts, the scattered
villages were drawn into the vortex of exchange, and a rudi-
mentary division of labour was established between the village
and the town. Among the raw materials which flowed to the
towns were metals. Some of these, such as gold and silver,
were used for luxury articles, but others, especially copper,
and above all copper alloyed with tin, replaced wood and stone
in tool-making, and so revolutionised the handicrafts. The
new urban economy was based on bronze.
In Egypt there is only a single river, which floods the whole
valley regularly every year. This annual flood is the sole agent
for fertilising the soil. It was •therefore a matter of vital con-
cern to every farmer that he should receive from the flood
just enough water and no more — enough to fill his dikes but
not so much as to burst them; and of course he needed to be
warned in advance when the flood was due. It was therefore
necessary that the flood should be regulated throughout its
course from the head of the valley to the sea — a tremendous
INTRODUCTION
23
feat of organisation, demanding a highly-skilled service of
astronomers and agronomists such as could only be provided
by a central government. Hence the rapid consolidation of the
two kingdoms of Upper and Lower Egypt, which were united
shortly after 3000 B.c. under a single monarch. The Egyptian
Pharaoh owed his position to an economic need. He was the
head of a centralised state apparatus controlled by the priest-
hood.
Mesopotamia was not unified in this way, because the
agricultural conditions were different. There were two rivers,
served by several tributaries and interconnected by a network
of canals. The result was that the cultivated areas were less
interdependent. Here therefore the towns grew into autonom-
ous city-states, each with its own priesthood and its own
priest-king. The competition between them was intense, and
about 1700 B.c. the country was unified by force of arms under
the hegemony of Babylon.
Notwithstanding these differences, the class structure of
Egyptian and Mesopotamian society was fundamentally the
same. In both countries large-scale agriculture had developed
on the basis of what began as a new division of labour — a divi-
sion between the producers and the organisers of production.
The organisers were the priests. They provided the intellectual
workers — the astronomers, mathematicians, engineers, archie
tects, scribes — who were just as indispensable as the manual
workers. In time these custodians of the means of production
became owners. They used the authority derived from the
nature of their task to concentrate the surplus in their own
hands. This too was economically necessary for the develop-
ment of new techniques. Bronze-working, in particular, was a
complicated and costly process, impossible without capital.
And so the growth of the new economy had the effect of con-
solidating the state in the form of an absolute theocracy. In
Egypt, the whole country belonged to the god incarnate in the
king, and all the productive functions of society — husbandry,
handicrafts, exchange — were strictly controlled. In Mesopo-
tamia, each city constituted a divine household, owned by the
patron deity resident in its midst and administered for him
by his tenant, the priest-king. The strong collectivism of
24 STUDIES IN ANCIENT GREEK SOCIETY
these early city-states was a heritage from the neolithic village
community, just as the king and priests derived their authority
ultimately from the magical fraternities which had grown up
round the chieftaincy in the higher stages of tribal society;
but it was now fostered by the ruling class systematically as a
means of safeguarding their privileges. The rigid stratification
of society is seen in the city’s lay-out. In the centre, towering
over everything, stood the temple, large, luxurious, exquisitely
furnished, surrounded by offices, treasuries, granaries, ware-
houses, and workshops for the accommodation of officials,
craftsmen and manual workers of all kinds. Some of these
were slaves taken in war; others were free in name but
economically dependent on the priests, their masters, the
largest employers in the city. Outside lay the arable land. A
portion was let out to tenant farmers or worked directly for
the temple under some form of labour service. The rest was
divided into family holdings which were free of rent or other
formal obligations but subject to the moral exactions with
which a powerful priesthood always exploits the faith of the
masses. Only the pastures remained common.
It is important to remember, as Gordon Childe has pointed
out, that even the lowest-paid workers in Mesopotamia were
better off than the free and equal members of any neolithic
village. The urban revolution had brought about an absolute
rise in the standard of living. On the other hand, if we take
into account the enormous rise in the productivity of labour,
it is clear that relatively they were worse off. The gains won
from the revolution were unequally distributed. It was this
factor that eventually brought the expansion of the new
economy to a stop. While the ruling class devoted an increasing
portion of the surplus to luxuries, the masses of the people,'
whose purchasing power was arbitrarily restricted, went short
of many things that had come to be regarded as necessities.
Meanwhile the city-states were entering into competition
with one another for raw materials and markets, with the
result that the ruling class was only able to maintain its
standards by intensifying its exploitation of the primary pro-
ducers. From this contradiction there was no escape. Commer-
cial rivalries precipitated wars, waged with bronze weapons
INTRODUCTION
25
and ambitions aims, until the whole country was forcibly
brought together under a series of empires, in which the class
struggle, sharper than ever, was fought out in new forms and
;on a vaster scale.
In Egypt, shut in by deserts and short of shipbuilding
timber, there was less foreign trade, and so the exploitation of
the primary producers was more intensive and direct. The
peasants were conscripted in masse to build for their rulers
sumptuous tombs, which, since they were places of worship
requiring priests for their maintenance, were a source of
revenue for the living as well as a memorial to the dead. Forced
labour and extortionate tribute reduced the mass of the popula-
tion to a condition little better than slavery. At the same time
the monarchy was faced with opposition from the more
powerful nobles, who tried to shake off the burden of royal
taxation and set themselves up as independent rulers on their
own estates. About 2200 B.C. the Old Kingdom collapsed
in civil war, but the paramount need for a central government
reasserted itself, and the monarchy was restored. The Pharaohs
of the Middle Kingdom pursued a policy of cautious expan-
sion, trading and raiding as far north as Syria, and so prepared
the way for the full-blown imperialism of the XVIUth
Dynasty. The stage was thus set for a conflict of empires.
The Babylonian Empire fell and was succeeded by the Assyrian,
the Assyrian by the Persian, the Persian by the Macedonian.
The Assyrians, Persians and Macedonians all conquered
Egypt, to be followed in their turn by the Romans and the
Arabs. For over five thousand years, during which they have seen
many changes of masters, the peasants of the Nile have continued
in poverty and sickness to till the richest fields on earth.
It is characteristic of the urban revolution that the great
alluvial valleys, which could alone provide the surplus re-
quisite for extensive metal-working, are naturally deficient in
mineral wealth. The metals had to be imported: copper from
Iran, Armenia, Syria and Sinai; tin from Iran and Syria; gold
from Armenia and Nubia; silver and lead from Cappadocia.
Thus trade was the life-blood of the new economy, and, as
it expanded, it drew an ever wider circle of neolithic villages
and mountain tribes into the orbit of civilisation.
26 STUDIES IN ANCIENT GREEK SOCIETY
By about 3000 B.c. the use of copper had been diffused over
the whole of the Middle East, but it was far from universal.
Even in Mesopotamia the cost of bronze remained high, and in
Egypt throughout the Bronze Age the peasantry continued to
work with tools of wood and stone. In the more backward
areas only chiefs could afford the new metal, and they used it
for swords, not ploughshares. Even where it was plentiful, the
people seem to have found it more profitable to export it un-
wrought than to develop a local industry. Arid so the earliest
urban communities to spring up outside Mesopotamia and
Egypt were primarily trading settlements. In Cappadocia, for
example, Kanes was founded by Mesopotamian merchants en-
gaged in trade with the local tribes, including the Hittites, who
controlled the mines of Mount Tauros. Similarly, in the north
of Palestine, where there was plenty of excellent timber as well
as rich deposits of copper and tin, a cluster of towns, including
Byblos and Ugarit, built up a prosperous trade with Egypt and
later expanded into first-class city-states, handling a vast
amount of traffic between Egypt, Mesopotamia, and Anatolia.
The Mediterranean was now thrown open to the urban
revolution, with all the advantages of maritime transport. The
first merchants to sail from Ugarit were doubtless bound for
the Delta or Cyprus, the island of copper. The urban develop-
ment of that island was, it appears, retarded by its wealth of
copper. Being so close to the more advanced communities
on the Syrian coast, the islanders devoted their energies to
exporting the metal in ingots instead of developing an industry
of their own. In any case, lying against the rugged south coast
of Anatolia, Cyprus was not well placed for trade.
It was otherwise with Crete. Equidistant from, Syria and
Egypt, it lies across the entrance to the ABgean basin, that
extraordinary amphitheatre of islands and mountains, which
leads through landlocked bays and winding valleys into the
Balkan highlands and so on to the Danube and Ceritral
Europe. In the course of the fourth millennium neolithic
immigrants found their way tentatively into Thessaly and the
Peloponnese. The earliest known settlers in Crete were also
neolithic, coming partly from Anatolia and partly from the
Delta. They settled in the east and south. Meanwhile the use of
THE EASTERN MEDITERRANEAN „ , Map I
28 STUDIES IN ANCIENT GREEK SOCIETY
copper had penetrated through the heart of Anatolia to the
JE gean coast, followed by a gradual growth of population, and
about 3000 B.c. some of these people took to the sea and
settled in the Cyclades and Crete.
The agricultural resources of Crete were small compared
with those of Egypt or Mesopotamia. There wer-e good pastures
and several plains suitable for grain, vine, palm and olive, but
a great part of the island consisted of mountain and forest, and
of course the sea was a barrier to expansion. On the other hand,
abundance of timber and good harbours enabled the islanders
at a very early date to take advantage of their maritime
situation. The result was that the wealth of their towns was
predominantly commercial, and the rapid growth of trade
acted as a check on the concentration of power in the hands of
large landed proprietors. The typical Minoan town clustered
round an open space adjoining the palace of a prince, one who
was high priest as well as governor, but primarily a merchant
prince, with other merchants living close by in mansions only
less rich than his, and with nothing to segregate either him or
them from the rest of the community. The very planlessness of
these towns bears witness to the greater freedom and flexibility
of social relations; and this means that in Crete, as compared
with Mesopotamia and Egypt, the urban revolution had been
carried through with less disintegration in the tribal structure
of society.
During the Early Minoan period (2900-2200 b.c.), in
which the use of metals was introduced, the main direction of
trade was towards Egypt and the Cyclades, and urban develop-
ment was confined to the east and south. In- the Middle
Minoan period (2200-1600 b.c.), marked by the development .
of bronze, we observe a steady growth of population, intensi-
fied trade with Egypt and direct intercourse with Syria.
Some time after 1700 b.c., when the East was thrown into
disorder by the Kassite conquest of Babylon, communications
with Syria were interrupted, and the Minoan princes sought
new openings in the Aigean. They strengthened their relations
with the Cyclades and established settlements in the Argive
plain and Central Greece. These developments gave the lead
to Knossos. In the Late Minoan period (1600-1200 B.c./ the
INTRODUCTION
2 9
princes of Knossos consolidated their hold over the island by
constructing a network of roads guarded by forts and they
extended their empire overseas to the Cyclades, Argolis and
Attica, perhaps even to Sicily. Their power was broken about
1450 B.C., probably by Minoanised chieftains from the Greek
mainland,, who invaded Crete and burnt her cities to the ground.
The empire held together for a couple of centuries longer, with
its centre at Mycen®, which entered into direct relations with
Egypt and the Levant. Then it collapsed after barbarian hordes
had swarmed down into the AEgean and overrun the whole of
the Eastern Mediterranean by land and sea as far as the Nile
Delta.
Mycenaj was not a town of the Minoan type. Its nucleus was
a heavily fortified citadel. Here, well protected, stood the
palace and storehouses, surrounded by the dwellings of the
nobility. Below the citadel lay an open settlement of craftsmen
and traders who served the palace needs. The ruling dynasty had
risen to power by its monopoly of bronze, which it used
primarily for war. The other centres — Tiryns, Thebes, Troy —
conformed to the same type.
The supremacy of these Mycenean princes was shortlived.
They had won their way to power by applying the technical
achievements of Minoan culture to the art of war. In particular,
they introduced the horse and chariot, and new types of sword,
rapier, helmet and body-armour. They did nothing to improve
the technique of production. And so they succumbed to a
fresh wave of invaders, who, being armed with iron, proved
more than a match for the bronze-clad knights of Mycenas.
The Dorians owed their superiority, not merely to the tise of
iron, though it was cheaper than bronze, but to the fact that,
since they were still organised on a tribal basis, it was available
to the rank and file as well as to the leaders. It was not a class
monopoly. And so the end of the Bronze Age coincided with
important changes in the_structure of Greek society.
Part One
" KINSHIP
Many proofs might be given to show that the early
Greeks had a manner of life similar to that of bar-
barians to-day.
THUCYDIDES
I
TOTEMISM
I. The Comparative Study of Ethnology and Archeology
The tribal peoples that survive to-day have been assigned to
the following categories according to their mode of food-
production: Lower Hunters (food-gathering and hunting);
Higher Hunters (hunting and fishing); Pastoral (two grades);
Agricultural (three grades ). 1 The Higher Hunters are dis-
tinguished from the Lower by the use of the bow in addition to
the spear, the arts of pottery and weaving, and the domestica-
tion of animals. In the Second Pastoral grade cattle-raising is
supplemented by agriculture; in the Third Agricultural,
garden tillage, done with the hoc, is superseded by field tillage,
done with the plough, and agriculture is combined with
cattle-raising. In these two grades we find further progress in
the handicrafts, permanent settlements, intertribal barter,
and metallurgy. At this level the tribal structure of society,
inherited from the lower grades, is beginning to break up.
This classification is of course an abstraction. Since it deals
with an organic process, it cannot be anything else. The
.categories are not mutually exclusive. Hunting and even
food-gathering are maintained in the higher grades, but with
diminishing importance. Nor do they constitute a fixed
chronological sequence. Food-gathering and hunting have come
first everywhere, but the higher grades depend on the local
fauna and flora and other environmental factors. In many regions,
where the natural conditions are favourable, tillage and cattle-
raising have been combined from the beginning in the form of
pastoral husbandry or mixed farming.®
1 Hobhouse 16-29. t ^ ie sake of simplicity I have omitted the
grade of Dependent Hunters. Current views of totemism are surveyed by
Van Gennep EAPT. My own has been anticipated by A. C. Haddon: see
Howitt NTSEA (1904) 154, Russell 1. 96.
® Childe MMH 85, Heichelheim 1. 48.
C
34 STUDIES IN ANCIENT GREEK SOCIETY I
Turning to prehistoric archaeology, we find that the hunting
grades correspond approximately to the upper palaeolithic
epoch, and the remainder to the mesolithic and neolithic. The
successive phases of a neolithic economy may be illustrated
by a particular instance. The prehistoric culture of the Danube
basin is divided by archaeologists into three phases . 3 In Phase I
hunting is already subsidiary. There are small herds of swine,
sheep, and oxen, but the principal mode of subsistence is the
cultivation of barley, beans, peas, and lentils in garden plots
tilled with the hoe. There is a rude technique of hand-made
pottery and some knowledge of textiles. In Phases II and III
the handicrafts, improve and there is an extension of cattle-
raising due to increasing pressure on the cultivable soil.
These two fields of research, ethnology and archaeology, tell
us all we know about the prehistory of human society, but
they have not yet been effectively co-ordinatecf. That eth-
nological data can be of great assistance to the archaeologist
no one would deny. An example lies to hand in the Danubian
culture. The excavations show that, though these settlements
were distributed densely and uniformly over the whole area,
none of them was occupied for more than a brief space of time.
The explanation is supplied by conditions that still prevail in
parts of Africa. A settlement is made on arable land, and the
soil is cultivated until it becomes exhausted. The settlement
is then abandoned and the cultivators move on. This is migra-
tory agriculture.
Archaeology deals with the material remains of extinct com-
munities. It tells us nothing directly about social organisation,
and some authorities deny that this gap can be filled from our
knowledge of modern tribes subsisting at the same material
level. Are we to assume, Gordon Childe asks, that, ‘because
the economic and material culture of these tribes has been
arrested at a stage of development Europeans passed through
some ten thousand years ago, their mental development
stopped dead at the same point ?’ 4 To this question he returns,
quite rightly, an emphatic negative. But the problem cannot
be left there. If the two sets of data are comparable at all, as
admittedly they are, it is incumbent on us to work out the
a Childe DEC 96-108. 4 Childe MMH 51; see below n. 6i.
I
TOTEMISM
35
appropriate comparative method. This is a task as difficult as
it is important. All that can he done here is to lay down some
guiding principles.
Modem capitalist civilisation has grown out of the pre-
historic cultures of Europe and the Near East, which developed
with exceptional rapidity. In contrast to these, the primitive
cultures still surviving in other parts of the world are products
of retarded or arrested development. These are the two ex-
tremes, and before arguing from one to the other we must find
some means of analysing this complication. It is a problem of
uneven development.
As Gordon Childe remarks, the social institutions of these
modern tribes have not remained stationary. They have con-
tinued to develop, but only in directions determined by the
prevailing mode of production. This is the key to the problem.
If, for example, we examine the Australian forms of totemism,
exogamy, and initiation, and compare them with similar in-
stitutions elsewhere, we find that they are extraordinarily
elaborate, pointing to a long period of development. But these
are all institutions characteristic of a simple hunting economy.
In other words, just as the economic development of these
tribes is stunted, so their culture is ingrown. And con-
sequently, while we cannot expect to find such institutions in
palaeolithic Europe in the same form, we are likely to find
them there in some form.
Again, just because of their backwardness, these tribes have
been exposed over a prolonged period to the influence of other
more thriving cultures with which they have come in contact.
Cultural diffitsion has of course operated in all ages, but its
effects are cumulative, and in these modem tribes they have
been exceptionally protracted and intense. Here again the
Australians are an extreme case. While retaining their paleo-
lithic economy, they have been subjected in recent times to the
impact of European capitalism, which is rapidly exterminating
them. It must never be forgotten that the primitive peoples
surviving to-day are known to us only to the extent that they
have been penetrated by our own traders, missionaries,
government officials, and ethnologists. In some cases they have
been converted outright into proletarians, like the Bantus in
I
36 STUDIES IN ANCIENT GREEK SOCIETY
the South African goldfields; in others their native institu-
tions have been arbitrarily stabilised as an instrument of in-
direct rule by the British Colonial Office. Such cultures must
of necessity present special features .due to the abrupt nature
of their contacts — features which can only be explained after
a methodical analysis of the effects of capitalist exploitation.
And that is a task which no bourgeois ethnologist is prepared
to undertake.
With these reservations the comparative method is an in-
strument of which we can and must avail ourselves if we are
intent on the advancement of our subject. ‘It has been proved’,
as De Pradenne courageously declares, ‘that with a more
limited scope prehistory cannot make progress: it comes to a
dead end, it marks time, it sinks into quicksands. To attack
all the problems along the whole line is the only way to reach a
solution.’ 2 * * * 6 Nor can we wait till we have perfected our tools.
We can only improve them by using them. It is necessary
to face the risk of error in order to discover truth.
2. The Origin of Totemism
Totemism is the magico-religious system characteristic of
tribal society. Each clan of which the tribe is composed is
associated with some natural object, usually a plant or animal,
which is called its totem. The clansmen regard themselves as
akin to their totem species and descended from it. They are for-
bidden to eat it,® and perform an annual ceremony to increase its
numbers. Members of the same totem may not intermarry.
Totemism survives most completely among the lower
hunting tribes of Australia. It is also found in forms more or
less disintegrated in America, Africa, India, and other parts
of Asia,* 7 and the European, Semitic, and Chinese civilisations
contain numerous traditions which have been recognised either
6 Dc Pradenne 14.
6 The taboo is directed primarily against eating the species, not against
killing it: Spencer NTCA (1904) 149, Conversely, a man may not eat
the species of another dan without permission: ib. 1 59, 296, NTNT 324.
7 The question of Indo-European totemism is only touched on by
Frazer (TE 4. 12-4) and ignored by Lowie (131). On Semitic, Chinese, and
Indian totemism see Robertson Smith RS, Granct 180, Ehrenfels MRI.
i
TOTEhiisM
37
as actual survivals of totemism or as relics of the ideology,
tenacious because so deeply rooted, which totemic practices
have generated.
The importance of Australian totemism is that it represents
the most primitive stratum of which we have direct know-
ledge. If from an analysis of Australian totemism in its present
form we can deduce its original form, and relate both to a
coherent evolutionary process, the result may be accepted as an
approximation to die history of totemism in general.
The great majority of Australian totems are edible species of
plants and animals . 8 * The remainder are mosdy natural objects,
like stones and stars, or natural processes, like rain and wind.
These inorganic totems are secondary, formed by analogy on
the pre-existing pattern. In seeking the origin of the system
we must concentrate on the plants and animals, and the fact
that most of these are edible is a pretty broad hint that its *
origin is connected with the food-supply.
The ceremonies for the propagation of the totem species are
performed at the opening of die breeding season at a pre-
scribed spot, called the totem centre, on the hunting ground of
the clan to which the totem belongs. The totem centre is
usually an actual breeding place of the species in question.®
If we ask what brought the ancestors of, say, the witchetty-
grub clan to the spot where ceremonies for the propagation of
witchetty-grubs are now performed, the answer can only be
that they came there to eat witchetty-grubs.
At the present day the clansmen are forbidden to eat, though
not necessarily to kill, their totem species, but to this rule
there are significant exceptions. In Central Australia, at the
performance of the increase ceremony, the headman of the
clan is not only permitted but obliged to eat a litde of the
species. As he explains, he must ‘get the totem inside him’ in
order to work his magic . 10 This ritual infraction of the taboo
8 Out of 200 totem species enumerated by Spencer and Gillen over
150 are edible: NTCA (1904) 768-73.
8 Spencer NTCA (1904) 147, 288, Frazer T 59, 62, 69, 70, 99, 185,
189. The ceremony is held annually at the opening of the breeding season:
Frazer T 72, 78, 195.
10 Spencer NTCA (1904) 323, NTNT 198, A 82.
38 STUDIES IN ANCIENT GREEK SOCIETY ■ I
is derived from the general practice of earlier times. That is
proved by the tribal traditions, in which the clan ancestors
are represented as feeding habitually or even exclusively on
their totem species . 11 This shows that totemism goes back to a
time when the technique of hunting had been so rudimentary
as.to impose severe restrictions on the quest for food, resulting
in a specialised diet . 12 The totemic clan originated in a small
nomadic band or 'horde' attracted to the breeding ground of a
particular species of animal or plant, on which it fed. It
remains to be seen how this state of affairs was transformed
into its opposite.
The increase ceremony is designed to represent dramatically
the growth of the totem, if it is a plant, or, if it is an animal,
its distinctive habits, movements, and cries, and in some cases
the act of catching it and killing it. Usually it includes a dance
in which the performers, appropriately disguised, mimick the
species to perfection, and sometimes they make a drawing or
painting of it on the rocks or in the sand. The original object
of such performances was probably actual practice in the
behaviour of the species, whose habits had to be studied before
it could be caught. Later, with the improvement of technique,
this function was superseded by that of a magical rehearsal.
By mimicking in anticipation the successful operation of the
quest for food the clansmen evoked in themselves the concerted
energy requisite for the real task. This is the essence of magic.
Magic rests on the principle that by creating the illusion
that you control reality you can actually control it. It is an
illusory technique complementary to the deficiencies of the
real technique. Owing to the low level of production the
human consciousness is as yet imperfectly aware of the
objectivity of the external world, which accordingly it treats
as though it were changeable at will, and so the preliminary
rite is regarded as the cause of success in the real task; but at
the same time, as a guide to action, the ideology of magic
11 NTCA (1904) 321, 324, 394, 405, A 331-2, 334, 339, 341-2;
see my AA 419 n. 6.
12 1 do not mean that the totemic species became the sole diet, which
continued to depend mainly on food-gathering', but that it was the species
on which the hunters concentrated.
I
TOTEMISM
39
embodies the valuable truth that the external world can in
fact be changed by man's subjective attitude towards it. The
huntsmen whose energies have been stimulated and organised
by the mimetic dance are actually better huntsmen than they
were before.
The members of the clan have a strong sense of affinity, even
identity, with their totem species . 13 The men who live on
witchetty-grubs, thriving when they thrive, starving when they
starve, and dramatically impersonating them in order to con-
trol them, are literally flesh of their flesh and blood of their
blood — a relationship which they express by saying that they
are witchetty-grubs. Hence, when the authority exercised by
the clan elders gives rise to ancestor worship , 14 the ancestors
are not worshipped in human shape but in that of the totemic
animal or plant.
It appears, then, that the first stage in the evolution of
totemism was the segmentation of the primitive horde, which
divided in order to gain access to different sources of food-
supply. So long as the new groups thus created lost touch with
one another, the change was merely quantitative — two groups
instead of one; but at some stage it became qualitative. Instead
of continuing to get their food independently by simple ap-
propriation, they became integrated as a pair of interdependent
clans. The food produced by each was distributed between them,
and this system of co-operation was maintained by means of a
taboo on die direct appropriation of the totem species — that is to
say, it could not be eaten when and where it was found, but
had to be brought home to be distributed. Each group became a
totemic clan, sharing its products with the other clan. How
this interchange was effected will be discussed later.
As the mode of production improved, this system lost its
economic basis. The quest of witchetty-grubs being no longer
a specialised technique, the function of the witchetty-grub
clan became purely magical — to make the species increase and
multiply for the benefit of the community ; 1 5 and the taboo on
13 An Arunta man, pointing to a photograph of himself, said, ‘That
one is just the same as me — so is a kangaroo* (his totem}: Spencer A 8o.
14 Landtman 125.
18 Spencer NTCS (1904) 327.
40 STUDIES IN ANCIENT GREEK SOCIETY I
the totem species, being now cut off from its economic origin,
became absolute.
Meanwhile the ceremonies themselves were modified.
TWfad of representing the activities of the totem species as
such they became celebrations of events in the life, of the
totem ancestors. This too can be studied in Central Australia . 1 6 •
The ceremony is still regarded as necessary for the fertilisation
of the species, but that is now done through the agency of the
ancestors, whom the dance calls into action; and in this form
the ceremony serves the further purpose of transmitting the
clan traditions to the rising generation . 17 In this way a pro-
cedure which began as an inseparable part of the mode of pro-
duction is converted into a purely magico-religious system
providing a sanction for the social structure which has grown
out of it.
In Australia the ideology of totemism has been expanded
into a comprehensive theory of the natural world. Just as the
social organism consists of so many clans and groups of clans,
each with its own totem species, so the world of nature — the
sea, streams, hills, heavenly bodies, and all that dwell therein —
are classified on the totemic model. The various kinds of trees
are grouped with the kinds of bird that nest in them; water is
assigned to the same group as waterfowl and fish . 18 The world
of nature is reduced to order by projecting on to it the organisa-
tion imposed by nature on society. The world order is a re-
flection of the social order — a reflection which, owing to
man’s weakness in the face of nature, is still simple and
direct.
In other parts of the world, where economic progress was not
arrested at this early stage, the whole system has collapsed,
leaving only a sense of kinship inspired by common descent, a
distinctive ancestral cult, the practice of exogamy, a purely
formal taboo on a particular plant or animal, and a prolifera-
tion of totemic myths.
16 Spencer NTCA (1904) 297.
17 lb. 328-92, Landtman 21, 31, Webster 27, 32, 60, 140.
18 Howitt NTSEA (1904) 454, 471, Radcliffe-Brown SOAT 63, R. B.
Smyth 1. 91, Durkhcim FPC, Radin 141.
TOTEMISM
41
3* The Origin of Exogamy
Membership of the dan is determined by descent. In the last
century, following Bachofen, ethnologists were agreed that
descent was .reckoned originally through the mother. To-day
this view is rejected by nearly all authorities outside the Soviet
Union, but without any agreed alternative. It has recently been
reaffirmed by BriflFault, who, arguing from a vast amount of
material, in collecting which he has shown far more energy
than his opponents, has, in my opinion, proved that the old
view is correct.
Many instances are recorded from modern tribes of the
transition from matrilineal to patrilineal descent, none of the
reverse process. 1 ® In Australia, where the two modes are found
in almost equal proportions and often intermixed, the in-
cidence of patrilineal descent rises in proportion to the elabora-
tion of the system of exogamy — a system which has grown in
some areas within living memory; and there is other evidence
of recent changes in the status of women.® 0 Elsewhere the
transition is known to have been promoted by contact with
European culture. A Chocta Indian once told a missionary that
he wanted to become a United States citizen, because then his
heic would be his own son and not his sister’s son. 21 In Nigeria,
where the transition is quite recent, it is attributed by the
natives themselves to the influence of British magistrates,
who persistently place their own bourgeois value on the rela-
tion between father and son. 22
Reviewing the evidence as a whole, we find that matrilineal
descent preponderates slightly in the hunting-grades, but then
declines, rapidly in the pastoral grades, much more slowly in
the agricultural. 23 This shows that the mode of descent is cor-
related with the mode of production.
In the pre-hunting stage there was no production, only
19 Cf. Smith and Dale 1. 292, A good example of an Indian matriarchate
transformed quite recently by the introduction of money is given by
Ehrenfcls 62.
20 Spencer A 150, 167, 328, 340, 346.
21 Morgan AS 166.
22 Meek 49, .61.
23 Hobhouse 150-4.
42 STUDIES IN ANCIENT GREEK SOCIETY I
simple appropriation of seeds, fruits, and small animals, and
consequently there was no division of labour. With the in-
vention of the spear, however, hunting became the men’s task,
while the women continued the work of food-gathering. This
sexual division of labour is universal in hunting tribes , 24 being
due 'to the relative immobility of women during pregnancy
and lactation . 26
Hunting led to the domestication of animals. Instead of
being killed the game was brought home alive and kept.
Accordingly cattle-raising is almost everywhere men’s work . 26
On the other hand, food-gathering led to the cultivation of
seeds in' plots adjacent to the settlement, and so garden tillage
is women’s work . 27 Then, after the introduction of the cattle-
drawn plough, agriculture was transferred to the men . 26 In
parts of Africa, where the plough is only a recent acquisition,
the change-over can be seen taking place at the present
day . 29
These shifting tensions in the relations of the sexes to the
mode of production explain the rise of patrilineal descent. The
process began with hunting, and was intensified by cattle-
raising, but in the initial phase of agriculture it was reversed.
How, it has been asked, if descent was originally matrilineal,
has it come about that some of the most backward peoples
reckon through the father, while others, more advanced,
retain the older form? The answer is that the sexual division of
labour characteristic of a hunting economy is such as to impart
to that economy an inherent tendency to patrilineal descent.
The reason why so high a proportion of modern hunting
94 Malinowski FAA 275-83, Bancroft 1. 66, 13 1, 186, 196, 2x8,
242, 261-5, 340, Heichelheim 1. 14. The need for the men to travel un-
encumbered save for their weapons explains why the women carry the
baggage: Basedow 112, Roscoe B (1911) 23, Landtman 15.
26 Zuckcrmann xo.
26 Landtman xj.Westermarck ODMI 1. 634, 2. 273.
27 Hobhouse 22, Heichelheim 1. 14, cf. Held. Pont. HP. 23, Eus. PE. 9.
26 Lowie 71, 174, 184, Childe MMH 138.
29 Krige 190: ‘Nowadays this rule [that the soil is tilled by women]
has been relaxed considerably owing to the influence of European civilisation;
with the introduction of the plough, for which oxen are used, men have
come to do all ploughing, because women may not work with cattle’.
I
TOTEMISM
43
tribes are patrilineal is that their economic life has been
arrested at that level. Conversely, when we find, as we shall
find, that in the prehistory of civilised peoples matrilineal
descent persisted to a much higher stage than the ethnological
data might lead us to expect, the explanation is that these
peoples passed rapidly through hunting to agriculture.
Where the nineteenth-century authorities failed was in their
attempts to account for the origin of matrilineal descent. Morgan
argued that in the conditions of collective marriage, which he
postulated for the early stages of society, the children were
necessarily assigned to their mothers’ clan because their
paternity was unknown. But in these conditions no significance
was attached to individual parenthood at all . 30 It was the
progressive definition of individual parenthood, determined
by the growth of individual rights of property, that destroyed
collective marriage. Morgan's theory is therefore in need of
modification at this point.
In two widely separated Australian tribes, of which we
happen to be exceptionally well informed, we find an elaborate
code of regulations requiring the married men to hand over
the whole or the best part of their catch to their wives’
parents . 31 Similar rules are common in other parts of the
world . 32 They point to a state of society in which the men
went to live with the clan to which their wives belonged — a
matrilineal clan centred in the women.
Another Australian tribe, the Yukumbil, has a tradition to
the effect that in old times, when the men went hunting, they
used to take their wives and children with them, but later
they found it more convenient to leave the children behind in
the care of an old woman . 33 This is a remarkable folk-memory
of the division of labour that followed from the development of
hunting. When the first camp was formed, the women took
charge .of it. The clan was centred in the women, and the
children belonged to the clan in which they were born.
30 A native of New Britain once boasted of having three mothers, and
these likewise asserted, ‘All three of us bore him': Frazer TE I. 305.
31 Spencer A 491, Howitt NTSEA (1904) 756-66.
32 Haddon RCAE 5. 149-50, Briffault 1. 268-430.
33 Raddiffe-Brown TEA 403.
44 STUfilfiS Ift ANCIENT* Gfi-EfiK SCClEtV i
The primitive horde was of course necessarily endogamous.
This too is remembered in Australian tradition. The tribal
ancestors are represented as mating invariably with women of
their own totem . 34 I have argued that the transition from the
primitive horde to the tribe — the complex of exogamous clans
—was dictated by the advance from appropriation to^produc-
tion, and that the economic interdependence of the clans took
the form of a taboo on the totem species, which obliged each
cl an to share with the others the food it obtained on its own
hunting ground. But why did not these clans continue to in-
breed like the parent horde? We have seen reason to believe
that each clan subsisted originally on a specialised diet, that the
men went to live with the clan into which they married, and
that they surrendered the products of their labour to the
members of that clan. In these conditions the practice of
getting husbands from other clans enabled each to extend, its
diet by obtaining access to foods which it did not produce
itself. The initial function of exogamy was to circulate the
food-supply.
The tribe is a multicellular organism which was evolved
from the primitive horde on the basis of a division of labour
determined by the low level of production, effected through
the rule of exogamy, supplemented by mimetic magic, and
projected ideologically in the form of zoomorphic ancestor-
worship.
Among the Lower Hunting tribes these totemic institu-
tions, though they have developed right away from their
original economic function, still form a coherent system, as
stable and definite as the tribes themselves. But when, under a
pastoral or agricultural economy, the tribal structure decays,
totemic magic, with its dramatic and pictorial representations
of the sacred plant or animal, with its implicit theory of the
kinship of all forms of life, and its practical function of
bringing the external world under control, breaks up into a
multiplicity of collateral activities, which, nurtured in turn
by new divisions of labour following from further develop-
ment of the productive forces, emerge as the arts and sciences,
myths, religions, and philosophies. Springing as it does from
34 Spencer NTCA (1899) 415.
I
TOTEMISM
45
the very moment — the advance from appropriation to pro-
duction — at which man parted company with the animals, it is
the matrix of human culture.
4. The Totemic Cycle of Birth and Death
Totemism has also left its mark on the life-history of the
individual.
In the beginning all labour was collective. The individual
was incapable of survival except as a member of a group. The
reproduction of the group was inseparable from production of
the means of subsistence.
In hunting tribes, besides the sexual division of labour dis-
cussed above, the danspeople are graded as children, adults,
and elders. The children help the women in their food-gather-
ing; the men hunt; the elders direct and supervise. 36 The basis
of these age-grades is physiological. The young and old are
dependent on the adults for their food. In their simplest form,
therefore, they are anterior to hunting. Originally those who
were past work were left to die, but later the aged, whose long
experience made them the natural repositories of traditional
knowledge, acquired an economic value and so were able to
assert a prescriptive claim to the surplus product of the group.
Child-getting, on the other hand, was always as vital as
food-getting. The whole training of the young was concen-
trated on these two techniques; and, since the female part in
reproduction is at once more apparent and more difficult than
the male, the magic invented to assist it bore from the outset a
feminine stamp.
The transition from one grade to the next is effected by
rites of initiation. The most important of these is the one
performed at puberty, when the adolescent became a full
member of the group, trained for production and reproduction.
The significance of this crucial change — physical, mental,
social, economic — is expressed in primitive thought by die
idea that at initiation the individual dies and is bom again. 36
36 The best study of this subject is still Webster PSS. There is no
monograph on the initiation of women.
36 Cf. Curcau 167: 'The natives hold that every serious event in physical
life is equivalent to death followed by a resurrection’.
I
46 STUDIES IN ANCIENT GREEK SOCIETY
This is one of the basic concepts underlying the whole history
of religion, and so it is important to understand what it
means.
The new-born child is greeted as one of the clan ancestors
come to life again — as a reincarnation of the clan totem . 87
That is why all over the world it is or has been the custom to
name the child after one of its progenitors 38 -^, custom often
associated with the rule that the person whose name is chosen
must be deceased . 38 The name is a totemic symbol, and there-
fore magical. The reluctance of savages to reveal their names to
strangers is well known. They are totemic secrets . 40 These ideas
are so radical that even in our own family of languages a
common base underlies the original words for name, mark,
kinship, knowledge (Latin nomen , not a, gens, gnosco ). The name
and the mark are the same thing, expressing, the one in oral
37 Karsten 416: ‘When a child is bom, the life thus brought into being
is not a new life. ... It is simply one of the forefathers that reappears
in the new-born. And on the other hand, when an Indian dies, he does not
cease to exist. Death does not imply the extinction of life, only a transition
from one form of life to another.’
38 Frazer TE 2. 302, 453, 3. 298, Karsten 417, Krige74, Hollis MLF 305.
So in Greece: Daremberg-Saglio s.v. Nomen. Cf. Frazer GB-TPS 320-7.
Smith and Dale 2. 59: ‘To get a new nam6 is to be reborn, remade’.
Gronbech X. 258: ‘Name and fate interpenetrate; the name was a mighty
charm, because it carried the history not only of the bearer but of his
ancestors and of the whole dan', lb. 287: ‘When a new man came into
the family, the Norsemen said expressly, Our kinsman is born again —
so-and-so has come back: and they confirmed their saying by giving
the old name to the young one'. In Chinese ming 'destiny' is the same
word as ming ‘name’ (Granet 249). In Greece names were bestowed by the
Moirai (Nonn. D. 46. 73, cf. A .A. 686-90, Pi. O. xo. 49-55, and see below
p. 338). Hence the new name assumed at initiation (Webster 40, Van
Genncp RP 120); at marriage, which was originally inseparable from
initiation (Smith and Dale I. 369, Meek 384, Hollis MLF 303, cf. below
n. 51); at coronation, which is a specialised rite of initiation (Hocart 77-98,
Meek 133, cf. below p. 158); in time of sickness, to make the patient ‘a
new man' (Frazer TE 2. 534, Roscoe B, 1911, 64); and at purification for
homicide (Apld. 2. 4. 12), purification being a form of regeneration. A new
name is assumed at the profession of vows in the Christian Church, and the
significance of the Christian name is explained in the baptism service: ‘Give
thy Holy Spirit to this infont, that he may be bom again'.
30 Morgan AS 78, Hutton 237, Playfeir 100.
40 Frazer TE x. 196-7, 489.
I
TOTEMISM
and the other in visual form, the totem incarnate in the
bearer. The kinsman is known by the name he bears and the
sign he wears — by his totem.
Just as the ancestor is born again as an infant, so at puberty
the child dies as a child and is born again as a man or woman.
And the occasion is marked by giving him a new name. The
adult is transformed by the same means into an elder. This
second stage has been less persistent than the first, but it
survives extensively in the ritual of admission to the status
of medicine-man or magician, and again the novice receives
a new name . 41 Finally, at death the elder is numbered among
the totemic ancestors, the highest grade of all, from which in
due course he re-emerges to pass through the whole cycle
again. Birth is death and death is birth. They are complemen-
tary aspects of an eternal process of change.
The re-birth of the initiate is represented dramatically. The
ceremony is often highly realistic — a close mimicry of the act
of dying and being born from the womb; or the novice pretends
to be devoured and disgorged by a god or spirit . 48 In die higher
cultures it assumes a more attenuated form, such as the magic
sleep or dream, in which the novice is laid to rest as a child
and wakes as an adult , 43 or the custom of dressing the boy as a
girl or the girl as a boy , 44 on die principle that before acquiring
the new identity he must escape from the old. When the
candidates for initiation are taken away from the village, their
mothers mourn for them as dead, and when they return they
behave like infants as though unable to walk or speak or
recognise their kinsfolk.
Another widespread feature of the ritual is a surgical opera-
tion or amputation of some part of the body — penetration of
the hymeneal membrane, circumcision or subincision of the
prepuce, knocking out a tooth, amputating a finger, cutting
41 Howitt NTSEA (1904) 738, Van Genncp RP 89, Webster 174-5.
48 Webster 38. Hastings 7. 318: ‘In the profession of vows in use
among the Benedictines the novice is laid out on the ground between four
candles and covered with a winding-sheet, the service of the dead is per-
formed over his body, and the whole congregation chants the Mistnrc for
him.' On Greek initiation see my AA 97-129.
43 Frazer TE 3. 370-456, Webster 154.
44 Halliday H.
I
48 STUDIES IN ANCIENT GREEK SOCIETY
the hair . 46 None of these has any utilitarian value except the
first, and it has been suggested that circumcision was modelled
in the first instance on the rite of perforating the hymen . 46
In all cases the amputated part is as a rule carefully preserved, 4 ’
and so these procedures present a parallel to the ritual of the
dead, whose bodies are preserved, in whole or in part, so' that
they may be born again. The same principle underlies the
worldwide practice of interring the corpse in the so-called
contracted attitude — arms and legs doubled up against the
chest — which reproduces the posture of the unborn child . 48
The remaining ceremonies consist of purifications and ordeals.
The novices are washed in water or blood, they bathe in a
running stream or the sea, or are scorched in front of a fire; they
run races, sometimes with painful handicaps, or engage in
sham fights, often with fatal results; they are scourged till they
swoon, their ears or noses are bored, their flesh gashed and
tattooed. The physical pain incidental to most of these ordeals
is everywhere explained as a trial of strength, in which failure
means disqualification and disgrace . 40 In many cases their
severity has been deliberately accentuated by the elders in
charge of the ceremony, who seek to terrify the novices into a
habit of unquestioning obedience ; 60 but behind them' all
lies the motive of mortification or purification, fertilisation or
regeneration. Just as pollution is disease and disease is death,
so purification is renewal of life.
Finally, the novices receive instruction in sexual and social
behaviour. This is done by homilies, catechisms, dramatic
46 Webster 32-8.
46 Briflault 3. 325-33.
47 Webster 36.
48 In modern tribes: Karsten 34-5, Krige 161, Junod SAT 1. 135 (cf. 166),
Earthly 78, 156, Smith and Dale 2. 104, Roscoe BB (1923) 292, BTUP
144, 154, 179, 198. In pre-dynastic Egypt: CAH X. 240. Sumer: CAH
377> Neolithic Europe: Burkitt P. 163, Chiide DEC Index s.v. Burials.
Neolithic Greece: Payne AG 150, Xanthoudides 134, Mylonas 424,
Frodin 433. Earthy 78 says expressly diat ‘the idea is to place the child
under similar conditions and in the same position as those in which it is
born.'
49 Webster 34-5.
60 lb. 59-66.
I
TOTEMISM
. 49
dances, and the revelation of sacred objects, especially symbols
of the sexual act . 61 The whole ceremony is secret. It is per-
formed at a distance from the settlement, usually on a prepared
ceremonial ground, from' which all save the elders and their
initiated assistants are warned away, often oh pain of death.
The actual initiation is frequently preceded by a probationary
period of seclusion, and when released the initiates are strictly
forbidden to divulge to the uninitiated anything they have
done or .heard or seen.
5 . From Totemism to Religion
Totemism differs from mature religion in that no prayers
are used, only commands. The worshippers impose their will
on the totem by the compelling force of magic , 62 and this
principle of collective compulsion corresponds to a state of
society in' which the community is supreme over each and all
of its members. So long as the united efforts of the whole
community are absorbed in maintaining it at the bare level
of subsistence, there can be no economic or social inequality
beyond the prestige earned by individual merit . 68 This is still
the- case in Australia. The status of the Australian headman
.depends on general consent. There are no chiefs in the Aus-
tralian tribes and no gods.
The more advanced forms of worship, characteristic.of what
we call religion, presuppose surplus production, which makes
it possible for a few to live on the labour of the many. The
headmanship ceases' to be elective and becomes a hereditary
81 It. 49-58; see below pp. 241-2. Among most hunting tribes initiation
is followed immediately by marriage, which accordingly is not marked
by a distinctive ritual. The initiatory ordeals ' of young men are often
treated as a prerequisite for marriage, and sometimes inflicted by men of the
bride’s clan. Hence the worldwide institution of the pre-nuptial contest:
Briflault 2. 199-208, cf. Od. 21, E. Hip. 545 sch., FHG 2. 238, Parth. 6,
Hdt. 6. 126-30, Pi. O. 1. 69-89, Paus. 3. 12. 1.
82 Frazer T 257.
83 The status of elders in the Lower Hunting tribes is well illustrated
by Hose 2. 182. Spencer A 9: 'Old age does not by itself confer any
distinction, but only when combined with special ability; there is no such
thing as a chief of the tribe.'
D
50 STUDIES IN ANCIENT GREEK SOCIETY- I
chieftaincy. The totem is attended with prayer and propitia-
tion, assumes human shape, and becomes a god . 54 The god is
to the community at large what the chief is to his subjects.
He is endowed with all the qualities attributed to the ideal
chief and worshipped with ceremonies modelled on the service
of the real chief . 65 As a Greek proverb says, gifts move kings
and gifts move gods . 60 The idea of godhead is a projection of
the reality of kingship. In the human consciousness, however,
this relationship is inverted. The king is believed to derive
his power from God and his will is accepted as the will
of God.
The further expansion of class privilege fosters an increasing
complexity in the divine powers from. which it draws its
sanction. As the ruling clan extends its authority, it annexes
the totem gods of other clans and absorbs them into its own.
The royal totem becomes the god of the tribe or league of
tribes, and eventually of the state. Some gods are conquered
by others; wars between kings and nations are waged again in
heaven. The array of totemic emblems that made up the
regalia of the Egyptian Pharaohs symbolises the fusion -of
tribes which led to the unification of the kingdom, and the
ceaseless rivalries between the cities of the Tigris and Euphrates
are mirrored in the composite and unstable Babylonian
pantheon . 67
Yet these gods never shook off entirely the marks of their
64 NTCA (1904) 490-1, Howitt NTSEA (1904) 488-508. The first
stage in the evolution of an anthropomorphic deity may be studied in
Howitt’s account of the spirit Biamban, who was simply a projection of the .
natives’ ideal of a headman (506-7). A good example of the transition from
ancestral spirit to nature god is given by Junod 2. 324-5. On various at-
tempts to show that the Australians believed in God before contact with
missionaries see Spencer A 589-96, Briffault 2. 698-9.
66 Meek 217: ‘The workaday religion of the Jukun is the cult of ancestors;
on the national side this assumes the form of the cult of dead kings,
who become gods. Cf. 159: ‘The shrine of the god Adang is a replica
in miniature of the private enclosure of the chief . . . Rites which are the
counterpart of those carried out thrice daily for the living chief are performed
by the priest of Adang.’
60 PI. R. 390c.
67 Frazer TE 1. 8i, 2. 139, 151, x66, cf. 18, Moret 143-5, Robertson
Smith RS 73, Engels LF 65-9.
I
TOTEMISM
51
origin. They can still incarnate themselves in their animal
form; they still have their sacred animals, which appear as
their attendants or emblems ; 88 they are begotten by animals in
miraculous births. Religious symbolism is still permeated
with reminiscences of the animal origin of die godhead.
As the totem became a god, the totemic rite became a
sacrifice. In most pastoral communities the catde are used for
milk, not meat, and so the flesh, especially of females, is
tabooed . 80 The totemic taboo was thus adapted to a new func-
tion. And meanwhile the increase ceremony had become the
common meal at which die clansmen reunited from time to
time under the presidency of their chief to partake sacrament-
ally of the flesh of their sacred herds. The meal began with a
sacrifice — diat is, die first helping was offered to the clan god,
who ate with diem because he was their kinsman and enjoyed
precedence over their chief because he was their chief of chiefs.
Similarly in agricultural communities the offering of the
firstffuits to the chief or priest, representing the god, is a
survival from the time when the chief had been presented
with die first portion at the distribution of the crop . 00 Later
still the same pattern can be discerned in the ritual of mystic.il
brotherhoods. Under die direction of their priest men whom
the class struggle had humbled and oppressed ate the flesh and
drank the blood of dieir god, feeding on the illusion of a lost
equality. The belief that the god must die in order diat his
people might live was already implicit in die totemic rite,
in which the sacred animal was killed year by year to make it
multiply. Just as the sacrament is descended from the ritual
infraction of die totemic taboo, so the rite of communion is a
sublimated image of the communal consumption of die
wealdi produced by the communal labour of die clan.
88 The associated object is regarded in the first instance as a repository
of the divine energy: Karstcn 207.
80 Robertson Smith RS 223, Roscoc BB (1923) 6. Krige 55. This
rule is not universal: Hutton 69, Gurdon 51. Later still the slaughter of
plough-oxen is tabooed: Acl. VH 5. 14 (see below p. 122 n. 90).
«o Robertson Smith RS 244-54, cf. Junod SAT I. 595, 2. 10.
I
52 STUDIES IN ANCIENT GREEK SOCIETY
6. Totemistn in Paleolithic Europe
Most contemporary archaeologists reject the comparative
method.
We shall frequently invoke the ideas and practices of contemporary
savages to illustrate how ancient peoples, known only to archaeology, may
have done things or interpreted them. But save in so far as such modern
practice and belief are used as a mere gloss or commentary on actually
observed ancient objects, constructions, or operations, the usage is illegi-
timate. The thoughts and beliefs of prehistoric men have perished irrevoc-
ably save in so far as they were expressed in actions the results of which'
were durable and can be recovered by the archaeologist's spade, o 1
This grants both too much and too little. On the one hand, we
are not entitled to use ethnological data even as a gloss or
commentary until we have analysed and classified their social
context. We cannot assume, for example, that Bantu ideas of
the after-life are relevant to the interpretation of Aurignacian
interments, because Bantu society belongs to a more advanced
stage than Aurignacian. On the other hand, it is almost
meaningless to say that the thoughts and beliefs of prehistoric
man have perished save in so far as they are recoverable by
excavation. The whole question is how far. And there is only
one way of answering it — by considering the nature of primitive
thought in general, that is, by applying the comparative
method. If the problem is approached from this angle — if the
ground is properly prepared — we shall find that the archaeo- .
logist’s spade goes deeper than is usually supposed.
Among the palaeolithic remains thrown up by this spade
are the bones of dogs. These animals must have reacted to .
their environment in the same way as Pavlov's, because they
belong to the same species. Animal behaviour is determined
by the operation of physical impulses in response to external
stimuli. In man, however, these impulses have been modified
by social tradition, and to an increasing degree in propor-
tion as he has become civilised. Further, the development of
man s social tradition is determined by his use of tools — by
01 u C ^^ e MMH 53. Childe has revised his attitude, cf. AA 243:
.Sy an d anthropology . . , axe two complementary developments
of the science of man ... as mutually indispensable as palaeontology and
zoology in the science of life.’
I
TOTEMISM
53
production. The rich individuality of civilised thought, the
complexity of our social relations, the multiple divisions of
labour, the elaborate technique of modern industry — all these
are manifestations at different levels of the high development
of the productive forces, in virtue of which the human
consciousness has continuously extended its control of its
environment. As we descend the scale, the technique of pro-
duction declines, divisions of labour disappear, social organisa-
tion becomes simpler, the human consciousness more uniform,
more immediately determined by the mere struggle for exist-
ence, until we reach the level of the animals. To quote again
from De Pradenne, ‘the more primitive the stage of man s
development, the more closely is his life conditioned by his
environment ’. 83 This is just as true of palaeolithic man as it is
of the modern Australians. And in these two cases the mode of
production, dependent on food-gathering and hunting, is the
same. The comparability of the two cultures is thus proved by
their common economic basis.
It must of course be granted that all attempts to reconstruct
prehistoric culture are limited by what the spade reveals. But
what does the spade reveal?
The Australians are in the habit of decorating rocks and
caves with figures of men and animals, drawn or painted . 83
These ‘picture caves’, as they are. called, have been found as
far apart as Western Australia, the Northern Territory, and
Queensland. At North Kimberley, where they are specially
abundant, there appears to be one on the hunting ground of
each local group. The human figures are of both sexes, the
females with exaggerated sex marks. The animals and plants,
so far as they have been identified, are all edible species —
kangaroos, lizards, nalgo fruits . 84 There are also composite
designs, such as a man carrying a kangaroo, and a group of
female kangaroos with cubs in their pouches. Another com-
mon figure is the impress or stencil of a human hand, pro-
duced by smearing the inside of the hand with wet paint or
powdering the back of it after it has been laid on the rock . 86
For the interpretation of these designs we can appeal to the
62 D e Pradenne 12. 84 Elkin 277.
83 Grey 1. 201-6, Elkin 257-79. 86 Grey 204, Elkin 261.
54 STUDIES IN ANCIENT GREEK SOCIETY I .
natives themselves, who still use them for ceremonial purposes.
At the opening of the breeding season the pictures are re-
painted or touched up in order to bring rain or to propagate
the species represented. Abundance of kangaroos and nalgo
fruits is thus ensured, and women are made prolific . 60 This is
only another form of the increase ceremony. The aft of painting
is emerging into independence, but it is still tied to magic.
The technique is crude and probably decadent. Many of
the paintings are very difficult to get at. There is a cave at
North Kimberley with decorations on the roof, which can
only be seen by crawling a long distance on all fours and then
turning over on your back . 67 This suggests that in former
times, before the native culture was broken up, the ritual
was more elaborate.
Cave painting is not confined to Australia. The African
Bushmen, another totemic hunting people, have dwindled to
a few thousand stragglers, all in South Africa, but at one time
they must have roamed the whole continent, because their
pictures have also been found in the Sahara, In-Guezzam, and
the region of Lake Tanganyika . 68 The art is now dead, but it
was still living in the Transvaal fifty years ago, and the
natives are still able to explain it. It is superior in technique to
the Australian, and bolder in conception. One of the finest
examples portrays a herd of ostriches, one of which carries a
bow and arrows and walks on human legs . 60 This must be a
huntsman who has disguised himself to get within bowshot:
was he a member of an ostrich dan? In another we see half a
dozen men dancing. They are surrounded by onlookers of'
both sexes, who are dapping, and they are wearing antelopes’
heads . 76 This can only be the mimetic dance of an antelope
dan.
With this Bushman art we may compare the cave paintings of
upper palaeolithic France and especially eastern Spain . 71 The
06 Elkin 261-3.
67 Elkin 258.
68 Leakey 137-60, Burkitt SAP 110-59, Adam 85-92.
60 Adam 88.
76 Adam 4, cf. Schapera 203.
71 Burkitt P 192-221, Macalister 1. 455-505, Adam 69-77.
TOTEMISM
55
resemblance is so close that some authorities regard them as
the work of the same people. The palaeolithic subjects include
simple frets, spirals, and crude animal figures; these are suc-
ceeded by astonishingly lifelike stags, bison, and other animals,
hunting scenes, fighting scenes, and men wearing stags’ heads.
Another common design is the stencilled outline of a human
hand. 72 The caves show no sign of regular habitation, and
some of the paintings are even more inaccessible than those at
North Kimberley. The cave at Niaux, for example, is a mile
long. It has plenty of suitable surfaces near the entrance but
no traces of decoration for more than 500 yards. All archaeo-
logists are now agreed that die primary intention of these
paintings was magical.
There is of course an inherent difficulty in distinguishing a
man disguised as an animal /*•
from the animal itself, but \ Ji
some of the instances are
unmistakable. One of die
Pyrenean caves contains a
figure of a man wearing
stag’s horns and a short
tail. 73 In the rock shelter at
M&ge a stag’s horn was found
decorated with three human
figures dressed in chamois
skins, masked with chamois
heads, and poised as though dancing. 74 This is another
totemic dance.
These palaeolithic communities were totemic. That being so,
we must presume that they were acquainted with the totemic
cycle of birth and death. And here again the spade comes to
our assistance. Burial in the contracted position — the ‘uterine’
posture — is not found in Australia, but among more advanced
tribes it is common in all continents. Similarly, it does not
FIG. I. Chamois dance: paleolithic
stag’s horn
72 Macalister 1. 456. The hand-outline also occurs in Libyan caves: Peel 399.
73 Burkitt P 3 11, Baldwin Brown 123-4.
74 Burkitt P 308: fig. 1. Macalister’s objections to the totemic interpretation
of these paintings (1. 505) are due to misapprehensions of totemism.
56 STUDIES IN ANCIENT GREEK SOCIETY I
occur in palaeolithic interments but is almost universal in die
neolithic . 76 .
The characteristic Australian forms of initiatory mutilation
are subincision and tooth-evulsion. "Whether the former was
practised in palaeolithic Europe is a question the archaeologist’s
spade can never answer, but among the remains of- the North
African Capsian culture are a number of skulls with the upper
middle incisors missing. There is no doubt that they were,
removed artificially . 76 Here we have a palaeolithic rite of
initiation.
The sign of the outspread hand is still common as an
apotropaic symbol in the Mediterranean and the Near East,
where it may be seen imprinted on doors and walls and
tattooed on women's faces . 77 In several of the palaeolithic
examples one or more fingers are partly or wholly missing . 78
This is another initiatory mutilation, and among the peoples
practising it are the Australians and the Bushmen . 79 A
custom so remarkable in itself cannot have arisen for more than
one reason.
Lastly, if these prehistoric cultures were totemic, they must
also have been exogamous, because exogamy is inherent in the
structure of the totemic clan. The parallel is complete.
Archeology and ethnology concur in confirming the thesis
laid down by Morgan seventy years ago, that the tribal system
has been universally the initial stage in the social evolution of
mankind.
The archaeological data, which were unknown to Morgan,
are not in dispute; yet, despite his lead, they have been left
uninterpreted. The spadework has been done, and with con-
summate skill. Why then, with the material in their hands,
have his successors been so slow to put two and two together?
The reason is that they have lost his grasp of the unity and
continuity of human progress. It has become a point of
78 Burkitt P 163; see above p. 48 n. 48.
76 De Pradenne 16 1.
77 Macalistcr 1. 509, cf. Seyrig 189-92.
78 Macalistcr 1. 458, 511.
79 Howitt NTSEA (1904) 746-7, Krige 4; sec further Frazer FOT 3.
198-241, Luquett MD.
I
TOTEMISM
57
honour with bourgeois specialists in the social sciences not to
trespass on one another’s preserves. For arch apologists to avail
themselves of ethnological data, except casually and un-
methodically, is ‘illegitimate’. It is equally illegitimate for
ethnologists and social anthropologists to note the bearing of
their results on archaeology except as an incidental ‘curiosity’.
One of them writes:
What the anthropologist deals with is not the past but the present. . . .
That some of the beliefs and customs thus revealed and described are
.curiously like those of very early man buried in the remote past and perhaps
like those of our own forgotten ancestors, is another story. 80
So the ethnologists treat prehistoric totemism as the archaeo-
logists treat totemism in general. In both cases it is ‘another
story', which nobody is left to tell. To tell the whole story
from beginning to end would not only reveal the present as a
continuation of the past — it would lift the veil on the future.
There’s the rub.
80 Goldenweiser 47.
II
THE NOMENCLATURE OF KINSHIP
I. Structure of the Tribe
The primitive horde evolved by self-division. First, it split in
two; then each half split again into two or more units. This
gave a tribe of two moieties, each containing so many clans.
Then these clans divide, giving a tribe of two moieties in each
of which there are so many phratries or groups of clans. The
basic unit is the clan. The phratry is a group of clans evolved
from a single clan. The moiety is a group of phratries derived
from the initial bisection . 1 The tribe is the whole complex,
preserving the unity of the original nucleus.
In reality, of course, the tribal system did not develop with
such perfect precision. There were complications and devia-
tions. That was inevitable in an organic process operating in
different environments. Nor was it maintained simply and
solely by economic forces. A structure so delicate must often
have been mutilated by war and famine, and in particular
cases we know that it was reconstituted artificially by in-
corporating new clans from outside or by transferring old
clans from one phratry to another. But such arbitrary readjust-
ments are a testimony in themselves to the vitality of the
system and the strength of its hold on the human mind.
We are now in a position to formulate more precisely the
rule of exogamy. The rule applies to all sexual intercourse, not
merely to marriage . 2 In Africa and America the prohibition is
generally confined to marriage within the clan, but there is
evidence from North America that the exogamous unit was
1 The moiety survives as a functional unit chiefly in Australia, but
it can be traced all over the world: Spencer A 41-3, Rivers HMS 2.
500-6, KSO 205-6, Layard 53-73, Morgan AS 90-3, 166-7,
Dorsey 230-2, Radin 121, 141-2, 163, 265, Eggan 268, 287, Hutton 125,
Haeckel TZ, Frazer TE 1, 256-71, 3 14-514, 2. 274, 3. 33, 90, 119, 121,
125, 130, 266, 280.
a Hollis NLF 6, Roscoe BTUP 33, Hutton 133, Gurdon 194.
II ^ THE NOMENCLATURE OF KINSHIP 59
formerly the phratry, 3 "and in the more backward Australian
tribes it is still the moiety. 4 The exogamy of the phratry
dates from the time when that unit had been a single clan, and
the exogamy of the moiety takes us back to the origin of the
rule in the initial bisection of the horde.
The moieties mark the decisive step in the construction of
the system — decisive because it was die first. The continuous
intermarriage characteristic of a tribe divided into exogamous
moieties produces automatically an intricate network of
relationships in which each individual is bound to all the
others by a double tie of blood and marriage. These interrela-
tionships are reflected in the nomenclature of kinship, which
is designed to express them. And the nomenclature tends to
persist after the actual relationships on which it rests have
been modified. Hence the study of primitive terminologies of
kinship provides a clue to the prehistory of marriage.
It is a fundamental postulate of historical linguistics, to
which this study belongs, that words change more slowly
than the meanings attached to them. An examination of
these terminologies shows in almost every case discrepancies
between the relationships actually existing and those implied
by the nomenclature of kinship, and discrepancies of this
kind are evidence that the nomenclature has been inherited
from an anterior stage in which it corresponded to the
reality. This principle was enunciated by Morgan at a
time when both sciences, linguistics and ethnology, were
in their infancy, and the whole study of evolution, physical
and social, has proved that it is correct. Just as biology,
the study of the structure of extant living organisms, is
reinforced by palaeontology, the study of fossils, so by
applying the linguistic method to primitive peoples, whose
history may be otherwise unknown, we can penetrate their
past.
Starting from these premisses, let us review the three main
types of kinship' terminology distinguished by Morgan. His
results were based on an analysis of 150 languages from all
continents except Australia. I have collected and analysed
s Morgan AS 90, Frazer TE 3. 79, cf. Buradkar COG
4 Frazer TE 1. 339-95.
60 STUDIES IN ANCIENT GREEK SOCIETY II
about 130 more, including those now .available from Australia. 8
My work on this subject has convinced me that his general
conclusions are sound, but I have amplified them at some
points, particularly in regard to certain deviations from type,
which he did not know of or did not explain.
2. The Classificatory System: Type I
The first type is found in a number of Polynesian languages
and in one Australian, the second in Australia, Polynesia,
India, North America, and parts of Africa. These are the two
forms of what Morgan called the classificatory system. The
third, called the descriptive system, occurs sporadically in
Asia and America, notably among the Eskimos, but with
these exceptions it is confined to the Indo-European and
Semitic languages.
Type I is very simple. There are only one or two terms for
each generation. All members of my own generation are my
‘brothers' or ‘sisters’ — that is to say, the terms applied to the
actual brother or sister are also applied to all cousins to an
infinitely remote degree. Similarly, in the first ascending
generation all are ‘fathers’ or ‘mothers’; in the first descending
generation all are ‘sons’ or ‘daughters’ or in some languages
just ‘children’ without distinction of sex. For the second
ascending and descending generations there is only a single
term of common gender comprising both grandparents and
grandchildren together with all their collaterals.
8 My chief sources, apart from Morgan, are as follows. Australia: Spencer
NTCA (1899) 66, 77, 79 . NTCA (1904) 77-8, 80-8, NTNT
65-81, A 41-61, Howitt NTSEA (1904) 160, 169, NTSEA (JAI) 287,
Radcliffe-Brown SO AT, Cameron 354. Oceania: Hose 80, Seligman
MBNG 66, 481, 707, Rivers HMS 1. 28-32, 177-92, 2x4-93, 2 99 > 34 1 *
376, 392-8, 2. 506, Codrington 35, Ivens IBP 76, C. E. Fox 20, Firth 248,
Layard 127-32, Fortune 37, G. Bateson 280, Williamson 2. 148, 198;
201-12. Africa: Seligman PTNS 52, 1 17, 152, 218, 258, 315, 379, 434,
507, Roscoe B (1911) 130, BB (1923) 18, NB 273, 292, Torday 285,
Rattray TTA 1. 1-41. America: Eggan SANAT, Rojas KN, Hoebel CS,
Strauss SU. Asia: Czaplicka 30, 35, 41, 59, Man 421, RaddifFe-Brown
AI 54-6, Seligman V 64, Pcrera 143, Rivers T 483, Hutton 139, Mills
AN 128, 163, RN 128, I. H. N. Evans 303, Lin Yueh-Hwa KSL,
Sturtevant HG. Europe: Kretschmer GBB, Durham 151-2.
II THE NOMENCLATURE OF KINSHIP 6l
Morgan argued that this type points ro a time when there
was no restriction on sexual intercourse within each genera-
tion. My father may be my mother's brother, my mother may
be my father’s sister; my brothers and sisters arc identical
with my brothers-in-law and sisters-in-law, and their children
arc indistinguishable from my own children. This is the
endogamy of the primitive horde.
The use of a single common term for die two outlying
generations reflects the division of the group into the three age-
grades — immature, adult, and senile. Each child, as it learns
to speak, finds itself in the lowest grade of a community
divided into 'grandparents', 'fathers’ or ‘mothers', and ‘brothers'
or ‘sisters'. At puberty the child enters the second grade, and
thereupon a new grade emerges of 'sons’ or daughters’, but
meanwhile die 'grandparents’ have disappeared. 0
- There arc two terms for ‘brodicr’ and ‘sister’, one of them
applied by a man to his brodicrs and by a woman to her
sisters, the other applied by a man to his sisters and by a
woman to her brodicrs. Thus, in Tikopia taina means ‘brother’
when die speaker is a man, 'sister' when the speaker is a
woman; have means ‘brother’ when die speaker is a woman,
'sister' when the speaker is a man. These terms arc called ‘self-
reciprocal’. If A is taina to B, B is taina to A. The common
term for ‘grandparents' and 'grandchildren' is also self-
reciprocal. Thus, in Dobu my grandparents arc tuhuna to me,
and I am tuhuna to them. This principle is a radical feature of
die classificatory system. In some Polynesian languages it can
even be traced in the terms for parents and children. For
example, tama, the proto-Polynesian word for 'father', means
in some languages ‘son’ or 'daughter'. In Tikopia we have
tama, 'son' or ‘daughter’, by the side of tamana, 'father'.
It seems probable that the whole system was originally self-
reciprocal. This would give us an original set of dircc terms,
0 It is a general rule — in Australia almost universal — that the elder
brother and sister, together with their classificatory collaterals, are dis-
tinguished from die younger by separate terms. This is the only age dis-
tinction characteristic of the system, and I agree with Krichcvsky 257-328
that it is not original, being probably based on seniority in respect of
initiation (Rivers KSO 187-9).
Table I
Actual Relationship
Father's father
Mother's father
Father’s mother
Mother s mother
Father
Father's brother
Mother's sister's husband
Father’s sister's husband
Mother's brother
Father-in-law
Mother
Mother s sister
Father’s brother’s wife
Mother's brother’s wife
Father s sister
Mother-in-law
Brother
Father’s brother s son
Mother s sister s son
Mother's brother's son
Father's sister s son
Brother-in-law
Sister
Father’s brother s daughter
Mother s brother s daughter
Father s sister s daughter
Sister-in-law
Son
Brother’s son (m.s.)
Sister's son (w.s.)
Brother's son (w.s.)
Sister s son (m.s.)
Son-in-law
Daughter
Brother s daughter (m.s.)
Sister s daughter (w.s.)
Brother s daughter (w.s.)
Sister’s daughter (m.s.)
Daughter-in-law
Son’s son
Daughter’s son
Son s daughter
Daughter s daughter
II I in
Tikopia
tubuna tupuna
tamana tamana
bwosianaj tamana fongovai
smana tinana
masikitang;
lawana | tinana fongovai
/tasinaN /taina\
\nuuna/ \kave )
ma, taina
/nuuna\ /have \
\tasina/ ^tainaj
natuna tamafine
kedeana
tubuna makopuna
TERMINOLOGIES OF KINSHIP
6 3
Telugu
BEtlSflH
Actual Relationship
tata
grandfather
Father's father
thunthi
Mother’s father
nowillie
avva
grandmother
Father's mother
kadnini
Mother’s mother
nia
tandri
Father
unde
Father’s brother
Mother’s sister's husband
H
mama •
mena-mama
mama
Father's sister’s husband
Mother's brother
father-in-law
Father-in-law
. luka
talli
[mother
Mother
aunt
Mother’s sister
Father’s brother’s wife
nowillie
atta
men-atta
atta
Mother’s brother’s wife
Father’s sister
mother-in-law
Mother-in-law
/nuthie \
-Vkupukay
/anna \
\tammudu/
brother
Brother
cousin
Father’s brother’s son
Mother's sister's son
wittewa
bava
Mother's brother’s son
Father’s sister’s son
brother-in-law
Brother-in-law
/kakua \
\kupuka/
/akka \
Vchelleluj
sister
Sister
cousin
Father's brother’s daughter
nupa
vadine
Mother's brother’s daughter
Father’s sister's daughter
Sister-in-law
biaka * *
koduku
son
Son
!
nephew
Brother’s son (m.s.)
Sister’s son (w.s.)
thidnurra
men-alludu
alludu
Brother’s son (w.s.)
Sister’s son (m.s.)
son-in-law
Son-in-law
biaka
kuthuru
daughter
Daughter
!
niece
Brother's daughter (m.s.)
Sister’s daughter (w.s.)
thidnurra
mena-kodalu
kodalu
Brother’s daughter (w.s.)
Sister's daughter (m.s.)
Daughter-in-law
Son s son
Daughter's son
, , , Son s daughter
manamaralu granddaughter — — — = :
* • ® Daughter s daughter
manamadu
II
64 STUDIES IN ANCIENT GREEK SOCIETY
Explanation of Talk 1
The first column gives a list of relationships extending as far as the first
collateral line of descent. Abbreviations: m.s. man speaking, w.s. woman
speaking.
Columns I-ffl show how these relationships are classified in Types I and
II of the classificatory system and (III) in the descriptive system. The
categories are delimited by horizontal lines.
The remaining columns give the actual terminology of five languages. .
Some details have been omitted. Dobu (Polynesia) conforms to Type I
except in having separate terms for the parents-in-law, brother's children
(w.s.) and sister’s children (m.s.), which are all developments in the direc-
tion of Type II. Tikopia (Polynesia) is intermediate between Types I and II.
The duplicate terms for brother and sister in these two languages are used
according to the sex of the speaker. Urabunna (S. Australia) and Telugu
(S. India) belong to Type II. The duplicate terms for brother and sister in
these two languages are used to distinguish the elder and younger in relation
to the speaker. As applied to the ortho-cousins, they distinguish the father's
elder brothers’ children from the father's younger brothers' children, and
the mother’s elder sisters' children from the mother's younger sisters’
children. The descriptive system is exemplified by English.
Sources: Fortune 37, Firth 248 and Rivers HMS 1. 299, 341, Spencer and
Gillen NTCA (1899) 66, Morgan SCA 523 no. 2.
one of which was used between alternate generations, the
second between adjacent generations, and the third within the
same generation. And the three terms would correspond to the
different modes of behaviour characteristic of the three age-
grades.
Morgan’s interpretation of Type I was challenged by Rivers
on the ground that the most primitive type of the classificatory
system is not likely to have been preserved by the Polynesians
to the exclusion of other more backward peoples. His own
view was that the Polynesian terminologies of this type are
degenerate. The distinctions lacking in these languages, as
compared with Type II, have been lost. This is not borne out
by the internal evidence, so far as it has been collated. The
Polynesian words for ‘mother’s brother’ or ‘father’s sister’,
where they exist, are either isolated forms, confined to one
language or locality and therefore not referable to the proto-
Polynesian system, or else they are compounds based on the
primary words for ‘father’, ‘mother’, ‘brother’, ‘sister’,
which are distributed with remarkable uniformity over the
II THE NOMENCLATURE OF KINSHIP 65
whole region : 7 Fotuna tua-tina 'mother’s brother’, from tua
‘brother’ and tina ‘mother’, Nokanoka nganei-tama 'father’s
sister’, from ngane ‘sister’ and tama ‘father ’. 8 The Tonga word
for the mother’s brother's son or daughter is compounded in
the same way from the primary terms for these three rela-
tionships (tama-a-tuasina), while the corresponding word in
Fiji (tavale) means literally concumbens , 8 and is therefore
properly an epithet of the primary terms for brother and
sister. If these compounds are secondary, as they clearly must
be, so are the distinctions they serve to mark.
It is true that Polynesian society is in many respects ad-
vanced, but it had no metallurgy. Tins should be considered in
conjunction with another circumstance. The Polynesian area,
which consists of a multitude of small islands scattered over a
vast expanse of the Pacific, is the most uniform linguistic
domain in the world. The Polynesians colonised it between the
tenth and fourteenth centuries a.d. — a navigational feat which
shows that their culture was then more advanced than it is
now. In other words, after reaching the zenith marked by the
period of migrations, their culture stagnated. That explains
why their languages have suffered so little change during the
7 The wide range of the primary Polynesian terms can be seen from
the following examples: tama 'father' occurs in this form in Motu,
Trobriand, Tube-tube, New Ireland, Bugotu, Florida, Eddystone, Guadal-
canal Pentecost, Fiji, Samoa, and cf. tamana (Tikopia, Aniwa, Fotuna,
Dobu), taman (Kayan), tamau (Kingsmill), tamai (Mota, Tonga), sama (Duff,
cf. sina ‘mother’), etma (Anaiteum, cf. etpo ‘grandparent’), timin (Weasisi),
rimini (Kwamera, cf. rini ‘mother’), ta (Tavua, cf. Navatusila ngwani-ta
‘father’s sister'), ama (Nokanoka, cf. ina ‘mother’), amai 'father's brother’
(Kayan), ma (Nggao, Loh, Narambula), maa (Lau, Fiu), man (Savo, cf.
Arosi matt ‘mother’s brother’), wama (Rafurafu, cf. waforo), mama (Koita,
Vella Lavella, Hiw, cf. Rafurafu tnamau 'mother’s brother'), imam (Vanua
Lava, Rowa), ma-hta (Hawaii) etc.; tina 'mother' occurs in this form in the
Solomons and Fiji, and cf. tinana (Tikopia, Fotuna), tinan (Kayan), sina
(Motu, Tube-tube, Duff), sinana (Dobu), tinau (Kingsmill), rini (Kwamera),
etna (New Ireland, cf. Anaiteum etma ‘father’), ina (Nokanoka), etc.
8 The form ngane appears to be derived by procope from *tua-kanc
(Samoa tua-ngane, Duff to-kane, etc.), the second element denoring the sex
either of the relative or of the speaker: hence Tavua ngwandi (* ngwane-tina )
mother's brother.’ Cf. also Mota ra-veve ‘father’s sister’ from veve 'mother'
with honorific prefix ra.
8 B. H. Thomson 371.
E
II
66 STUDIES IN ANCIENT GREEK SOCIETY
period of their separate existence. If it should turn out that
the migrations were preceded by a very rapid advance from
primitive beginnings, the anomalous survival of a primitive'
type of kinship system would fall into place as part of a process
marked by exceptionally sharp dialectical contradictions. •
3. Ritual Promiscuity
The primitive horde has disappeared from the face of the
earth, and so those of Morgan's opponents whose sense of
human dignity has been affronted may take comfort from the
thought that direct evidence for sexual promiscuity is neces-
sarily lacking. But, as we have learnt from totemism, social
institutions rendered obsolete by economic progress find a
sanctuary in religion, which is of interest to the historian of
humanity just because it is a stratified repository of discarded
practices and discredited beliefs. Long after men have ceased
in normal life to do as their forefathers did, they cling to the
belief that their prosperity depends in some way on the good-
will of their ancestors, and consequently, at critical moments
in the life of the individual or in times of public calamity,
ancestral customs tend to be revived.
In the Arunta tribe of Central Australia every woman is
required before marriage to have intercourse in a prescribed
order with several men who stand to her in certain prescribed
relationships, all of which except the last fall within the
prohibited degrees. 10 The act of marriage is preceded by a
formal acknowledgment of the wider ancient rights.
In the same tribe, and in many others, every married woman
is required once in her life to attend a ceremony in which she
is treated for the time being as the common property of all the
men present without regard to the rules of exogamy except
that her father, brothers, and sons are excluded. The natives
say that the licentious character of these occasions conforms to
the practice of their ancestors. 11
In the Fiji Islands, when a chief falls ill, his son presents
himself to a priest with a request to be initiated in order that '
his father may recover. The novice dies so that the sick man
10 Spencer A -472-6. 11 Spencer A 472-6, NTCA (1904) 73.
II THE NOMENCLATURE OF KINSHIP 67
may live. After the initiation a public festival takes place at
which all rules of exogamy and rights of property are sus-
pended. 'While it lasts,' a native blandly remarked, 'we are
just like the pigs.' Brothers and sisters, who in ordinary life
are forbidden even to touch one another, behave as man and
wife. The double significance of this ceremonial reversion to
communism, sexual and economic, is aptly expressed in the
native saying that on these occasions 'there are no owners of
pigs or women'. The details were recorded by Fison, a devout
but honest Christian missionary, who said:
We cannot for a moment believe that it is a mere licentious outbreak
without^ an underlying meaning and purpose. It is part of a religious rite,
and is supposed to be acceptable to the ancestors. But why should it be
acceptable to them unless it were in accord with their own practice in the
far-away past? 12
4. The Classijicatory System: Type II
In Type IT of the classificatory system each category of
Type I is bisected. There is a separate term for the mother's
brother as distinct from the father and father’s brother, and
this term includes the father-in-law. There is a separate term
for the father's sister as distinct from the mother and mother's
sister, and this term includes the mother-in-law. There are
separate terms for the children of the mother’s brother and
father’s sister as distinct from the brother and sister, who are
still equated with the children of the father's brother and
mother’s sister, and these terms include the brother-in-law and
sister-in-law. The terms for the son and daughter are applied
by a man to his own children and his brother's children, by a
woman to her own children and her sister’s children, but there
are separate terms for a man’s sister’s children and a woman’s
brother’s children, and these include the son-in-law and
daughter-in-law. The father’s parents are distinguished from
the mother’s, and the son’s children from the daughter's.
As in Type I, each term is used in the classificatory sense,
that is, it covers an infinite series of collaterals. For example,
the term for the father includes the father’s brother, the
father’s father’s brother's son, the father’s father’s father’s
12 Fison 30.
II
68 STUDIES IN ANCIENT GREEK SOCIETY
brother's son's son, and so on; the term for the mother includes
the mother’s sister, the mother's mother's sister’s daughter,
the mother's mother's mother's sister's daughter’s daughter,
and so on. The term for the father’s sister includes the father’s
father's brother’s daughter, and the term for the mother’s
brother includes the mother's mother's sister's son. Similarly
the terms for the brother and sister include the children of all
those called 'father' or 'mother'; the terms for the son and
daughter are extended by a man to the children of all those
whom he calls 'brother' and by a woman to the children of all
those whom she calls 'sister'.
We see that the speaker's generation falls into two categories.
The first includes die brother and sister, the father’s brother's
children, and the mother's sister’s children. These are the
'ortho-cousins’. The second includes the mother’s brother's
children and the father’s sister's children. These are the
'cross-cousins'. It is important to grasp this distinction.
The cross-cousins include the brother-in-law, if the speaker
is a man, or the sister-in-law, if the speaker is a woman. Now,
if a man’s male cross-cousin is his brother-in-law, his female
cross-cousin must be his wife; and if a woman's female cross-
cousin is her sister-in-law, her male cross-cousin must be
her husband.
In most languages the husband and wife are denoted by
special terms, which will be considered presendy, but in
Australia the term for cross-cousin includes the wife, if the
speaker is a man, and the husband, if the speaker is a woinan.
In other words, the children of the mother’s brother and
father’s sister stand to a man in the relation of brother-in-law
and wife, to a woman in the relation of husband and sister-in-
law. Similarly, in the preceding generation the father-in-law
is the mother’s brother,- the mother-in-law is the father's
sister; in the succeeding generation a man's son-in-law is his
sister s son and a woman's son-in-law is her brother's son.
The whole system turns on the continuous intermarriage of
cross-cousins.
Cross-cousin marriage is die form of marital relations that
results from die intermarriage in each generation of two
exogamous groups. All relatives are classified according as they
II THE NOMENCLATURE OF KINSHIP 69
belong to the speaker’s own group or to the other. Con-
sequently, just as Type I expresses the relationships charac-
teristic of the endogamous horde, so Type II corresponds to a
community of two exogamous moieties. The difference
between them, which is simply that Type II bisects each
category of Type I, follows from the bisection of the horde.
5. Group-marriage
On this interpretation, and on no other, the logic of the
system is' apparent. The linguistic evidence is so conclusive
that it would have to be accepted even if it were unsupported.
In fact, however, cross-cousin marriage is still the rule
throughout Australia, in parts of Polynesia and Melanesia,
among a number of Dravidian tribes in India, and in various
parts of North, Central, and South America and Africa. 13
Cross-cousin marriage may be individual or collective. Out-
side Australia it is to-day everywhere individual, save in so
far as a man who marries an eldest sister has a claim on the
younger ones too as they come of age. In these conditions the
terminology, which rests on the principle of collective rela-
tionships, is contradicted by the actual practice. But in parts
of Australia cross-cousin marriage is, or was till recently, col-
lective. A group of brothers are mated to a group of sisters. 14
Here the nomenclature corresponds to the reality. There can be
no doubt that this was once the case everywhere with Type II.
Just as the bisection of each category of Type I limited the
endogamy of the horde by the rule of exogamy, so the absence of
further distinctions within the new categories argues that
sexual relations were not subject to any closer restriction.
Marriage was collective. Indeed, at this stage it is scarcely
correct to speak of marriage at all, because, as will appear later,
formal marriage marks the definition of those individual rela-
tionships which eventually superseded the collective. 16 In each
generation the men of the one moiety were the mates, actual
or potential, of the women of the other.
13 Briffaulc x. 563—84.
14 Howitc NTSEA (1904) 173-87, Spencer NTCA (1904) 73, 95.
16 Briffaulc 2. 1-96.
70 STUDIES IN ANCIENT GREEK SOCIETY II
Morgan’s theory of group-marriage has been strenuously
and obstinately contested. It was published seventy, years ago,
yet it is still being denounced as vociferously as ever. It is a
wonder the old man had so much blood in him. Again and
again we have been assured that fresh evidence has rendered
his conclusions out of date. This attitude would be more im-
pressive if it were backed by a reassemblage of the data, but
apparently the evidence that damns Morgan is so vast that.it
cannot be collected. His corpus of 150 languages could be
doubled or trebled at the present day, but it has not been.
The additional materials lie scattered about in hundreds of
monographs and periodicals, and the standard collection is still
his Systems of Consanguinity and Affinity (1871). In regard to
actual marriage customs, as distinct from the terminologies,
only one methodical attempt has been made to bring his work
up to date — by Briffault; 10 and Briffault is one of his strongest
supporters, trenchantly exposing the unscientific reasoning of
his opponents and marshalling on his side a mass of concrete
data far more copious and complete than has ever been
adduced against him. In saying this I have not forgotten
Westermarck’s History of Human Marriage. The reader who has
any confidence in that work should consult Briffault. 17
Lowie, one of Morgan’s latter-day antagonists, observes
that his belief in social progress ‘was a natural accompaniment
of the belief in historical laws, especially when tinged with the
evolutionary optimism of the seventies’. 18 So Lowie does not
believe in historical laws. He admits that his own view of
history is unscientific. Why then does he ask us to believe it?
What he says here is of course quite true in the sense that
Morgan’s work, which has justly been compared with Dar-
win’s, 1 ® was an intellectual masterpiece of capitalism in its
prime. It is also true that Lowie's disbelief in social progress,
expressed in caustic aphorisms about ‘that planless hodge-
podge, that thing of shreds and patches, called civilisation’, 20
is an equally characteristic product of capitalism in decay.
10 614-781. 17 lb. 1. 764-5, 2. 16-64, etc * 18 Lowie 427.
19 Engels UFPS x 5. 20 Lowie 428.
II
THE NOMENCLATURE OF KINSHIP
71
6. Decay of the Classificatoty System
■ Since the starting-point of group-marriage was the bisection
of the horde, the collective character of the relationship must
at first have been complete, all the males of the one clan being
mated with all their female coevals of the other; but
when the two original clans had segmented into groups of
clans, or moieties, the range of sexual relations, though still
nominally coextensive with the moiety, waS in practice re-
stricted to one or other of its constituent clans. Instead of one
collective union there were several. The same process was
repeated when the clan became a phratry, until eventually the
rule of exogamy was concentrated in the individual clan. This
is the culminating point in the evolution of the tribal system,
which, starting from the undifferentiated horde, has now
become a complex of moieties, phratries, and clans.
After this point is passed, the gathering forces of economic
and social differentiation, which determined the growth of
the system, become disruptive. As the mode of production
becomes individualised, it is brought into conflict with the
collective organisation of the producers. Each producer
becomes more possessive as he becomes more self-sufficient.
And so collective marriage breaks down. Instead of a group of
brothers uniting with a group of sisters on equal terms, each
brother marries one or more sisters on his own, with the re-
servation that they shall be accessible to the others when he is
away from home. Later still, having established a prior claim
on the inheritance as senior member of the clan, the eldest
brother acquires a corresponding right to the whole group of
sisters, leaving only the reversion of them to his juniors after
his decease.
The marriage of a group of sisters to one man is the sororate;
the right to an elder brother’s widow or widows is the levirate. 21
These worldwide customs mark a unilateral development of
individual marriage in favour of the sex which is now playing
the dominant role in production. The converse of the sororate,
known as fraternal polyandry — a group of brothers married to
one woman — is much less common, because the social dominance
21 On the sororate and levirate sec BrifFault 1. 614-29, 766-81.
II
72 , STUDIES IN ANCIENT GREEK SOCIETY
of the female sex tends to go with the survival of common
ownership and hence of group-marriage in its unmodified
form . 22
Returning to Type IT, we observe that, from the moment
when the moiety ceases to be the basic exogamous unit, the
system contains a contradiction. Within each category there
has emerged in practice a distinction lacking in the nomen-
clature — between a man's actual brothers and sisters, born of
the same collective union as himself, and his classificatory
‘brothers' and ‘sisters', born of other such unions; between his
immediate ‘fathers' and ‘mothers', including his actual
parents, and his more distant ‘fathers' and ‘mothers’, with
whom he is only remotely connected.
This discrepancy was met by the use of descriptive epithets
— ‘near brothers’ and 'far-away brothers', ‘true brothers', and
so on. Epithets of this kind, designed to limit the primary
terms, are a widespread feature of the system . 23 They introduce
a new principle, because these new categories of ‘near brothers'
and ‘near fathers’ are restricted to a definite number of in-
dividuals. And even so they are only a makeshift.
With the assertion of individual marriage rights it became
expedient to distinguish the actual husband and wife from the
other cross-cousins, the actual parents from the other ‘fathers’
and ‘mothers’, the actual parents-in-law from the other
‘mother’s brothers’ and ‘father’s sisters’. The strain set up by
this innovation was naturally most acute at the point im-
mediately affected, and accordingly most languages, outside
Australia and parts of Melanesia, have evolved separate terms
for the husband and wife. The secondary origin of these terms is
betrayed in many cases by their still recognisable meaning —
‘man’, ‘woman’, ‘partner’, ‘couple’, ‘two-joined’, etc . 24 But,
once admitted, this descriptive principle asserted itself at all
the critical points until the new unit — the individual family —
as ib, i. 628.
23 Morgan SCA 523 nos. i, 2, 4, 5, 17, Spencer NTCA (1899) 79,
NTCA (1904) 78, 85, 88, A 47-55, Rivers HMS 1. 192, 237, 248, 266,
z 75 > 376, T 483, Seiigman V 64, PTNS 507, Hutton 139, C. E. Fox 20,
Roscoe B (1911) 130, Meek 114.
24 Morgan SCA 369.
II THE NOMENCLATURE OF KINSHIP 73
had been finally delimited. The collapse of the classificatory
system of relationship was thus brought about by the collapse
of the tribal system of society.
Before pursuing the details of this process let us see what
happened to the kinship terminologies of peoples whose
development was arrested at the tribal stage.
There are two main deviations from Type II. The first is
peculiar to Australia, where the tribal system remained intact.
In that continent we find in many languages a type of termin-
ology which is baffling in its complexity until we realise that it
has been formed from Type II in exactly the same way as
Type II was formed from Type I. Just as Type II bisected each
category of Type I, so Type Ila, as it may be called, bisects each
category of Type II. Just as Type II restricted promiscuity by the
rule of cross-cousin marriage, so Type Ila restricts cross-cousin
marriage by segregating certain cross-cousins as unmarriageable . 2 B
In all these tribes marriage is prohibited between cross-
cousins of the first degree, and the whole terminology has been
reconstructed accordingly. Instead of one category of cross-
cousins there are two, unmarriageable and marriageable. The
first includes the children of the mother's brother and father’s
sister together with all whom these call ‘brother’ or ‘sister’,
namely, the children of the mother’s mother's sister's son and
of the father’s father’s sister’s daughter, and so on. The second
includes the husband and wife, the brothers-in-law and
sisters-in-law, the children of the mother’s mother’s brother’s
daughter and of the father’s father’s sister’s son, and so on.
Instead of one term for the mother’s brother, the father-in-
law, and all whom these call ‘brothers', there are two, one
for the mother’s brother and his classificatory ‘brothers',
another for the father-in-law and his classificatory ‘brothers’.
The same subdivision appears in the tribal organisation itself.
Instead of the normal structure of moieties and phratries we
find that each moiety contains two phratries and each phratry two
subphratries . 26 This is only another expression of the marriage
26 G. Thomson AA 395. Systems analogous to Type Ila, and even more
elaborate, have been traced in parts of Melanesia: Layard 143-53.
20 This is what is known as the 'eight-class' system: Spencer NTCA
(1899) 77-9, NTCA (1904) 78-85, NTNT 73-5, A 41-6.
II
74 STUDIES IN ANCIENT GREEK SOCIETY
rule embodied in the kinship system. I must take a wife from
one particular subphratry in the opposite moiety to my own.
The members of that subphratry are, in my generation, the
marriageable cross-cousins as defined above.
The Arunta have introduced a further complication. Not
only am I forbidden to take a wife from the non-marriageable
category of cross-cousins — I may not even marry a woman of
the marriageable category if she belongs to the same local
group as myself. This restriction too is expressed in the terms
of kinship.
The reader may well ask whether in these circumstances it
is not difficult for an Arunta man to find a spouse at all. It
is, so difficult that the extinction of the tribe is being hastened
by its own marriage rules. This feature of Australian society is
pathological.
It may also be asked how these rude aborigines retain their
grasp of a nomenclature so elaborate that it gives us a headache
even to study it in a diagram. Here there is no difficulty.
Having no corn to measure or cattle to keep, these black-
fellows cannot count beyond five , 27 but they carry the facts of
kinship in their heads with a facility that makes the white man
seem stupid. Our terminology, on the other hand, is just as
perplexing to them as theirs to us. The reason why they have
encumbered their classificatory system with so many com-
plications is precisely that they have been incapable of the in-
tellectual revolution of thinking it out afresh in terms of
individual relationships.
Type Ha is everywhere associated with patrilineal descent, v
and it is reported to be still spreading at the present day . 28
These signs of recent growth enable us to explain it.
Backward though they are, these tribes have been in con-
tinuous contact for a century or more with European gold-
diggers, sheep-farmers, missionaries, policemen, and other
champions of our own culture. They have imbibed respect for
private property along with belief in God. By banning mar-
riage between cross-cousins of the first degree, and between
those belonging to the same local group, they have reduced to
a minimum the blood-bond between husband and wife, and
27 Spencer A 21. 28 Frazer T 5, 52, 256.
II THE NOMENCLATURE OF KINSHIP
thereby strengthened the husband's authority. As Spencer and
Gillen perceived, the special features of their kinship system
mark ‘the initial stage in die segregation of individuals to form
definite families in the sense of this term as used by us ’. 29 They
are an attempt to formulate a rule of individual marriage
within a system which being moribund is too rigid to be
radically reconstructed.
Similar factors have been at work among the Nordi American
Indians, who present the most characteristic examples of our
second deviation, Type lib. In die western and central States
the general rule is that a man must find a wife, not only out-
side his clan, but outside the first three degrees of collateral
descent — that is, a woman who stands beyond any effective
claim of consanguinity . 30 This too is probably a recent
development, because some of the tribes scill retain the simple
form of cross-cousin marriage . 31
Most of these Amerindians belong to die Higher Hunting
or First Agricultural grades. Their tribal institutions are
more advanced and consequently less stable than the Austra-
lian. In them, therefore, the effect of individual marriage
has not been to elaborate the classificatory system but to
dislocate it.
The weakest point in the system, after cross-cousin marriage
has been abandoned, is naturally the cross-cousin relationship.
Some means has to be found of distinguishing from die cross-
cousins the husband and wife and the brodier-in-law and
sister-in-law. Most of these languages have separate terms for
the husband and wife, diough several of them still include
the brother-in-law (woman speaking) widi die husband and
the sister-in-law (man speaking) with the wife . 32 Among the
Tinneh and the Rocky Mountain tribes the children of the
mother’s brother and father’s sister, being no longer marriage-
able, have been transferred to the category of brother and sister . 33
29 Spencer A 49.
30 Morgan SCA 164, AS 467.
31 Eggan 95, Brifiauk 1. 572.
32 Morgan SCA 291 nos. 26-7, 34-6, 53, Eggan 105.
33 Morgan SCA 291 nos. 56, 59, 63-4, 66. Possibly some of these
systems go back directly to Type I.
II
76 STUDIES IN ANCIENT GREEK SOCIETY
In Dakota they are designated by die terms for brother-
in-law (tahan, shechay ) and sister-in-law (hanka, echapan)
with the addition of a suffix (tahanshe, shechaysh } hankashe,
echapansht ). 34 Where one or other of these expedients has
been adopted, the terminology has remained stable. In a
great many languages, however, the cross-cousins have been
transferred to categories outside their own generation. In this
way a fresh contradiction has been introduced, into the system,'
leading in some cases to extraordinary confusion. Thus, in
Minnitaree, the mother's brother’s children are equated with
the son and daughter. Accordingly, their reciprocals, the
father’s sister’s children, are equated with the father and
mother, the father’s sister with the grandmother, and so on.
In Osage the reverse procedure has been adopted. The father’s
sister’s children are equated with the son and daughter, the
mother’s brother's children with the mother’s brother and the
mother. The further repercussions can be studied in Table
II . 3 6 All the dislocated Amerindian systems approximate to
one or other of these two types.
The reason why the terms for son and daughter have been
extended to the mother’s brother’s children in some languages,
and to the father’s sister’s children in others, is probably
connected with the sporadic practice of marriage with the
mother’s brother’s wife or the father’s sister’s husband . 36 In
the first case the mother’s brother’s children, in the second
the father’s sister’s, will be step-children, who in these languages
are commonly equated with the true children. Such marriages
are by their nature exceptional or occasional, and therefore
cannot have caused the dislocation, but they may have deter-
mined its direction.
This principle of consecutive dislocation is not confined to
34 Ih. 291 nos, 9-16.
38 Minnitaree type: Morgan SCA 291 nos. 26-32, 34-5, Eggan 289.
Osage type: Morgan SCA 291 nos. 18-24, 4$. 48, 52, 55, Eggan 252.
The two types are correlated with the mode of descent. In 8 out of 10 in-
stances of the former descent is matrilineal; in the other 2 the mode of
descent is not recorded. In 8- out of 12 instances of the latter it is patrilineal;
in 2 it is matrilineal. Sec below n. 37.
30 Eggan 274, Rivers HMS 1. 47-9, Junod LSAT 1. 266, 290, Earthy
14, Frazer TE 2. 387, 5x0.
II
THE NOMENCLATURE OF KINSHIP
77
Table II
DISLOCATIONS IN THE AMERINDIAN KINSHIP SYSTEMS
MINNITAREE
Actual Relationship
Equated with
Mother’s brother’s son
Son
daughter
son’s wife
Daughter
Daughter-in-law
daughter’s husband
Son-in-law
Father's sister's son
Father
daughter
Mother
son's wife
Mother
daughter’s husband
Father
Father's sister
Grandmother
Father’s sister's husband
Grandfather
Mother’s brother's son’s son
Grandson
son's daughter
Granddaughter
daughter’s son
Grandson
daughter’s daughter
Granddaughter
OSAGE
Actual Relationship
Equated with
Father’s sister’s son
Son
daughter
son’s wife
Daughter
Daughter-in-law
daughter's husband
Mother’s brother’s son
Son-in-law
Mother's brother
daughter
Mother
son’s wife
Mother’s bro's wi
daughter's husband
Father
Father’s sister’s son’s son
Grandson
son’s daughter
daughter’s son
Granddaughter
Grandson
daughter’s daughter
Granddaughter
America. It is also found in Melanesia and in Africa . 37 The
confusion to which it leads, especially in die relationship
between parents and children, shows diat the dassificatory
system has lost touch with reality. The new reproductive unit
is die individual family, comprising one man, one or more
sisters, and dieir offspring. The dassificatory system, designed
37 Minnicarce type: Rivers HMS z. 28, 30-1, 192 (all mntrilfncal).
Osage type: Roscoc BB (1923) 18, NB 292, Scligman PTNS 117, 258,
G. Bateson 280 (all patrilineal).
II
78 STUDIES IN ANCIENT GREEK SOCIETY
for an entirely different unit, is falling to pieces. The next
step, at which these tribes have stumbled, is to replace it with
a new system corresponding to the new reality.
7. The Descriptive System
The Indo-European family of languages is descended from
the speech of a people which occupied some part of the great
plain stretching eastwards from the Ukraine. Some time in the
third millennium B.C. this people broke up, migrating in all
directions, and their speech split into a number of derivative
languages, from which are descended the Indo-European
languages still living or preserved in written records.
Some archaeologists would identify the undivided Indo-
Europeans with the neolithic Kurgan culture of South Russia.
The tumuli or ‘kurgans’ from which this culture gets its name
have yielded pottery, horse-bits, and fragments of wheeled
waggons. This implies a nomadic pastoral economy with
access to forests. 38 The linguistic evidence indicates that when
they dispersed the Indo-Europeans were predominantly
pastoral with some knowledge of tillage and metallurgy; that
they were organised in clan settlements under some form of
chieftaincy or kingship; that descent was reckoned in the
male line; and that the women went to live with the clan or
household into which they married. 30 They -may accordingly
be assigned to the Second Pastoral grade.
Their primitive nomenclature of kinship has been recon-
structed by linguists, who had no knowledge of the classifica-
tory system, from a comparative analysis of the surviving
languages. It contains some apparent anomalies which they
have been unable to explain. 40 On the one hand, it appears to
have recognised no less than five, different relationships by
marriage; on the other, no primitive terms have been traced
for die mother’s brother, cousins, nephews and nieces, uncles
and aunts. At all these points it stands in striking contrast to
the later Indo-European terminologies and to the various
38 J. L. Myres in CAH 1. 83-5.
30 Childe A 78-93, MeillctlECLI 391.
40 Mcillct IECLI 389-92; see further my AA 402-17.
II . THE NOMENCLATURE OF KINSHIP 79
forms of the dassificatory system just reviewed in all parts of
the world.
- Of the surviving Indo-European terminologies die most
archaic is die Ladn. Let us sec what it contains.
In classical Latin dicrc are no specific terms for the diildren
of die fadicr's sister or modicr's brodicr, but the children of
my fadicr's brother arc my patrtteles and the children of my
mother’s sister arc my cotisobrini These are die ortho-cousins,
whom Type II of die dassificatory system equates with the
brother and sister. So in Latin: diese words arc properly
epithets of frater and soror, which indeed are frequendy ex-
pressed, c.g. f rater patruelis and jrater consobrinus as opposed to
frater gemanus 'true brodicr ’. 42 Furdier, die epithets can be
dispensed with. Frater and soror often stand alone for the
diildren of the father's brother or modier's sister : 43 that is to
say, they arc used in die dassificatory sense.
In Type H of the dassificatory system my father’s brother is
my ‘father’ and my modier's sister is my ‘modier’, but my
father’s sister and mother’s brother are denoted by different
terms. So in Ladn, my father's brodicr is my patruus , which
is merely an extension of pater, and my modier’s sister is my
matertera, an extension of mater, while my father’s sister is my
amita and my mother’s brodier is my atmculus.
Auonculus is a diminutive of auos, the Ladn for grandfather.
In the classificatory system die father's father is induded
under the same term as die modicr's mother's brodier. This is
because, with cross-cousin marriage, he is the mother's
mother's brother. If my modier's mother’s brother was my
auos, my own modier’s brodier might naturally be called my
auonculus.
Latin has lost the primitive IE terms for son and daughter.
41 The term consolrinus was sometimes • applied generally to any first
cousin (hence our 'cousin') but its original usage is fixed by its ety-
mology (*cousuesrinus), Matnulis for the mother’s brother’s son and amitinus
for the father’s sister’s son arc both late, being formed by analogy during the
codification of imperial Roman law.
42 Cic. Plane. 11. 27. Tin. 5. l. 1., Plaut. Aul. 2. 1. 3; cf. Irish dearbb-
bhr&tbair ‘brother,’ literally 'true brother,’ as opposed to br&thair 'brother
in religion’ (Old Irish bratbir 'brother' or 'father’s brother's son').
43 Cic. Clu. 24. 60, Att. l. 5. 1, Catull. 66. 22, Ov. Met. I. 351.
II
80 STUDIES IN ANCIENT GREEK SOCIETY
So has Celtic. As Vendryes has observed, this feature of the
Italo-Celtic group must have originated in some social change
that took place before the separation of Celtic from Italic . 44 The
Latin Jilius and j ilia are properly adjectives, referred conjec-
turally to /do 'suck '. 46 They are therefore analogous to patruelis
and consohrinus , which we have just recognised as descriptive
epithets of the classificatory terms.
As soon as we recognise the classificatory origin of the Indo-
European terminology, its anomalous features resolve them-
selves.
As a classificatory term, IE *auos had included the father’s
father and the mother’s mother’s brother. In Latin, Armenian,
and Old Horse it came to mean simply 'grandfather’; in the
Latin auonculus , Old Irish atnnair, Old High German oheim , and
Lithuanian avyttas, it was modified by an element-en affixed
to the stem and transferred to the mother’s brother . 46 In
French, Modern German, and Welsh the modified form has
been generalised as ‘uncle’.
The transference of *atios to the mother’s brother implies
the loss of an older term for that relationship. The lost term
was IE *suikuros, which had comprised the mother’s brother,
father-in-law, and father’s sister’s husband. This was ap-
propriated by the father-in-law (Latin sour ). IE *st}ikrus,
standing for the father’s sister, mother-in-law, and mother’s
brother’s wife, was appropriated in the same way by the
mother-in-law (Latin socrus ). Thus the term for the father’s
sister also disappeared. It was supplied in Latin by arnita ,
which is related to Old High German ana and Old Prussian
ane , both meaning ‘grandmother’. From this it appears that
the Latin amita, ‘father’s sister’, was formed by extension of the
stem from IE *ana denoting the mother’s mother and the
father’s father’s sister, just as the Latin auonculus, ‘mother’s
brother’, was formed from the IE term for the father’s father
and the mother’s mother’s brother.
The father’s brother and mother’s sister were distinguished
from the father and mother by extension of the stem.
Forms analogous to the Latin patruus and matertera exist in
44 Vendryes 26. 46 Walde-Pokorny 1. 830.
4 <s Emout-Meillet s.v. Avonculus.
lame ill
THE INDO-EUROPEAN NOMENCLATURE OF KINSHIP
Actual Relationship
11
Latin
Father's father
*auos
A
auos
Mother’s father
Father’s mother
auia
Mother’s mother
hbhi
Father
*patcr
pater
Father's brother
patruus
Mother's sister's husband
Father's sister's husband
*st^kuros
Mother's brother
auonculus
Father-in-law
socer
Mother
*mater
i
mater
Mother’s sister
matertera
Father's brother’s wife
Mother’s brother’s wife
*su£kriis
*
Father’s sister
amita
Mother-in-law
socrus
Brother
*bhrSter
frater
Father’s brother’s son
Mother's sister's son
Mother’s brother’s son
*daluer
ft
Father’s sister’s son
j
Brother-in-law
leuir
Sister
|
*su&or
A
t
soror
Father's brother’s daughter
Mother's sister’s daughter
Mother’s brother’s daughter
*g(e)l ou-
Father's sister’s daughter
Sister-in-law
glos
Son
*sunus
filius
Brother's son (man speaking)
nepos
Sister’s son (woman speaking)
Brother’s son (w.s.)
*geme-
Sister’s son (m.s.)
Son-in-law
Daughter
*dhughter
Bti
Brother’s daughter (m.s.)
nepos
Sister's daughter (w.s.)
Brother’s daughter (w.s.)
*snus6s
Sister's daughter (m.s.)
Daughter-in-law
nurus
. Son's son
*an£potios
nepos
Daughter’s son
Son’s daughter
*anepotia
Daughter’s daughter
II
82 STUDIES IN ANCIENT GREEK SOCIETY
Greek, Sanskrit, Old High German, Anglo-Saxon, and
Welsh. 47
IE *bhrater and *sufsor continued in Latin to include the
ortho-cousins. In Slavonic they were extended to the cross-
cousins. In the other languages — excepting Greek, which will
be discussed separately — they were restricted to the actual
brother and sister. The terms for the ortho-cousins were thus
lost.
IE *daiuer, comprising the brother-in-law and male cross-
cousins, was appropriated by the brother-in-law (Latin leuir).
The feminine *g(e)lou- was appropriated in the same way by
the sister-in-law (Latin glos). This removed the terms for
cross-cousins.
IE *sunus and *dhugbter were restricted to the actual son and
daughter, except in Italo-Celtic, where they disappeared. This
removed the designations for a man’s brother’s children and a
woman’s sister’s children. IE *geme~, which had comprised
the daughter’s husband, a man’s sister’s son, and a woman’s
brother’s son, and its feminine *snus6s, were restricted to the
daughter’s husband and son’s wife respectively (Latin gener
and mints). The terms for nephews and nieces were thus
eliminated.
We have seen that, with cross-cousin marriage, my* father’s
father is my mother’s mother’s brother. So, speaking as a man,
my son’s son is my sister’s daughter’s son. These are reciprocal
-relationships. Accordingly, just as IE *auos was divided
between the grandfather and the mother’s brother, the latter
being eventually generalised as ‘uncle’, so its reciprocal
*anfyotios was divided between the grandson and the sister’s
son, the latter being generalised as ‘nephew’. But, whereas the
second use of *auos was marked by modifying the stem, the
corresponding use of *anipotios was not, and consequently the
division was less definite. In Sanskrit it was restricted to the
grandson, in Old Irish to the sister’s son; in Greek, Old
Norse, Old High German, and Old Slavonic it was generalised
as ‘nephew’; in Latin, Old Lithuanian, and Anglo-Saxon it
fluctuated between the nephew and the grandson.
47 The Gk. tij'iTpu;, which is without parallel in the other languages,
was formed on the analogy of ir&Tpeos.
II THE NOMENCLATURE OF KINSHIP 83
There remains IE *jenater, denoting the husband's brother’s
wife . 48 This term is alien to the classificatory system, in which
the husband’s brother’s wife is identified with the sister. 4 ® It
belongs therefore to the last phase of the parent language, in
which, as we have seen, the social unit was the group of
brothers living with their wives, who came from other groups.
The Indo-European nomenclature thus falls into place as a
normal specimen of the classificatory system, Type II. It was
reconstructed by restricting each term to one of its several
applications, the nearer relationships being preferred to the
more remote and relationships dirough the husband to rela-
tionships through die wife. New terms were found for the
deprived categories by modifications of the stem, descriptive
epithets, and in some cases by transference to other genera-
tions. These are die same expedients that we have seen at work
in primitive languages all over the world. The Indo-European
system begins where the others leave off. If we put all the
evidence together, we cannot fail to recognise in it a single,
continuous historical process. In particular, the tendency we
have noted in the Indo-European system to distort terms by
extending diem beyond their proper generation confirms our
analysis of the more extensive dislocations characteristic of the
North American languages. And the reason why this tendency
was carried further in those languages than in Indo-European
is that the Amerindians have failed to advance beyond tribalism,
whereas the Indo-European-speaking peoples progressed so
rapidly that after only a brief period of instability their whole
system was reorganised on a new foundation.
This new foundation was the individual family. In the
descriptive system, the father is distinguished from his brothers,
the mother from her sisters, the brothers and sisters from
the ordio-cousins, the sons and daughters from the nephews
and nieces; the father-in-law and mother-in-law, brother-
in-law and sister-in-law, son-in-law and daughter-in-law, are
48 Lat. ianitriccs, Gk. dvtepEs, Sk. yatar, O. Si. jetry. Descriptive terms
for this relationship are not uncommon in classificatory systems: Morgan
SCA 291 nos. 3, 59, 63, 523 nos. 1-2, 4, 10, Seligman PTNS 2x8, 379,
Roscoe B (1911) 130 etc.
48 Morgan SCA 291 no. 64, Eggan 105 etc.
84 STUDIES IN ANCIENT GREEK SOCIETY II
also denoted by distinctive terms. The family is defined.
On the other hand, in contrast to Type II of the classifi-
catory system, the father's brother and sister are merged
with the mother's, the ortho-cousins with the cross-cousins,
the brother's children with the sister’s, the paternal grand-
parents with the maternal, the son’s children with the
daughter's. These distinctions, dictated by cross-cousin mar-
riage, have become superfluous.
When Morgan was working on these problems, the materials
for reconstructing the Indo-European nomenclature were
not available;- yet it was he who first drew attention to
the importance of the classificatory system for Indo-European
linguistics . 60 He saw that the descriptive system character-
istic of our languages could not be original. If subsequent
workers had followed his lead, the Indo-European nomen-
clature would have been explained long ago.
Morgan’s theory of the classificatory system was accepted by
the classical Australian field-anthropologists — Fison, Howitt,
Spencer and Gillen — who brought to light after his death new
data confirming his conclusions. Fison’s only hesitation was
prompted, as he candidly admitted in a letter to Morgan, by
the scandal caused among his religious colleagues . 61 Morgan
himself had trouble with the Rev. J. H. Mcllvaine, his local
minister, to whom he submitted the proofs of his Ancient
Society with the object of deleting anything that might be
judged incompatible with the Old Testament. When the
book was published, his spiritual adviser, blinded it seems by
affection for his friend, wrote to him: ‘I think it a great work,
and decidedly the strongest argument against the Darwinians
60 Morgan AS 491, cf. W. H. R. Rivers in Hastings 7. 703.
61 Fison wrote: ‘In my own mind I accept it [the Undivided Commune,
i.c. the endogamous horde] as sufficiently proved, but I do not positively
assert it for these two reasons: (x) I expect violent opposition and there-
fore resolved to narrow as far as possible the ground of controversy; (2)
the Undivided Commune means nothing more or less than "promiscuity”
and this would be terribly shocking to many of my best friends among our
nunisters. ... In short, I do not doubt the former existence of the Un-
divided Commune, but I do not consider it necessary for my purpose to
assert it, and moreover (owing to my surroundings) it were better for me not
to assert it so long as assertion is unnecessary’ (Stern 162). Life is thorny,
and whispering tongues can poison truth.
II THE NOMENCLATURE OP KINSHIP 85
and in favour of the permanent species that has ever been
given to the world ’. 62 .
Others were not so easily taken in. Marx immediately
acclaimed it, as he had acclaimed the Origin of Species at a time
when it was being indignantly denounced by the academic
world, and Engels declared that it 'has the same importance
for anthropology as Darwin's theory of evolution has for
biology and Marx’s theory of surplus value has for political
economy ’. 63 That of course is why, like them, it has been
condemned. The opposition to Darwin eventually collapsed,
because his theory was indispensable for industrial develop-
ment, but, outside the Soviet Union and the new democracies,
Morgan and Marx are still taboo.
There is more in it than religious prejudice. The family, as
well as God, goes hand in hand with private property. Ac-
cepting private property as something that 'was in the begin-
ning’, bourgeois thinkers have realised instinctively that
Morgan must be resisted all along the line. But, though
unanimous in opposing him, their front is not united, because
they have been totally incapable of finding an agreed alternative.
Radcliffe-Brown has argued that, ‘as against Morgan and
diose who follow him, it can be shown that there is a very
thorough functional relation between the kinship terminology
of any tribe and the social organisation as it exists at present’,
and hence ‘there is no reason whatever- to suppose that the
kinship terminology is a survival from some very different
form of social organisation in a purely hypothetical past ’. 64
The explanation of the classificatory system which he has con-
structed on these premisses is, as I have argued in detail
elsewhere, untenable . 66
62 Stem 27.
63 Engels UFPS 15.
64 -Radcliffe-Brown SOAT 427.
66 See my AA 396-401. Before challenging Howitt, Radcliffe-Brown, who
investigated the Karera at a time when they had dwindled to a few dozen
English-speaking stragglers hanging round the sheep-stations (TTWA 144),
might have heeded his warning (NTSE A, JAI, 278): ‘Unless an enquirer takes
note of the altered conditions in which the remnants of tribes are living . . .
his statements will conflict with those of earlier investigators who based their
views on the rules which obtained when the tribespeople lived a savage life.'
B6 STUDIES IN ANCiEtfT GREEK SOCIETY -II
Meanwhile another eminent anthropologist, Kroeber, has
been trying to prove the opposite. He denies that kinship
terminologies can be explained in the light of social organisa-
tion at all:
If it had been more clearly realised that terms of relationship are deter-
mined primarily by linguistic factors, and are only occasionally, and then
indirectly, related to social circumstances, it would probably long ago have
been generally realised that the difference between descriptive and clas-
sificatory systems is subjective and superficial. 69
The reader who has been cudgelling his brains to master the
Arunta system will be comforted to learn that objectively it is
the same as his own.
After this only one step was needed to remove the whole
problem from the realm of reality. This has been taken by.
Malinowski, who has discovered that ‘the plain fact is that
classificatory systems do not exist and never have existed ’. 67
Lowie has done the same with totemism. He is ‘not convinced
that all the acumen and erudition lavished on the subject has
established the reality of the totemic phenomenon ’. 68 The
problem is solved by denying its existence. This is the last
word in bourgeois scepticism, which, as usual, ends in flippancy.
Our relief at Malinowski’s discovery is a little dashed when
we find him confessing to the complete failure of the con-
temporary Anglo-American school of social anthropology to
perform their basic task:
As a member of the 'inner ring’, I may say that, whenever I meet Mrs
Seligman or Dr Lowie, or discuss matters with Radcliffe-Brown or Kroeber,
I become at once aware that my partner does not understand anything in the
matter, and I end usually with the feeling that this also applies to myself.
This refers to all our writings on kinship and is fully reciprocal. 69
So far are the doctors from agreement! After striving all these
years to refute Morgan they have only succeeded in refuting
one another. In the meantime Morgan’s work, as amplified by
Engels, is being carried on along a broad front by the eth-
nologists and archaeologists of the Soviet Union.
68 Kroeber 8a.
68 Lowie 137.
67 Malinowski K 22.
69 Malinowski K 21.
in
FROM TRIBE TO STATE
I. The League o f the Iroquois
Morgan's study of the Iroquois is a pioneer work of field
anthropology ana a masterpiece of its kind. It was during his
visits to these Indians that he found the clue to the tribal
organisation of ancient Greece and Rome.
In his general remarks on Amerindian society he says:
The plan of government of the American aborigines commenced with the
gens [dan] and ended with the confederacy, the latter being the highest
point to which their governmental institutions attained. It gave for the
organic series, first, the gens, a body of consanguinei having a common
gentile name; secondly, the phratry, an assemblage of related genres united
in a higher association for certain common objects; third, the tribe, an
assemblage of gences, usually organised in phratries, all the members of
which spoke the same dialect; and fourth, a confederacy of tribes, the mem-
bers of which respectively spoke dialects of the same stock language. It
resulted in a gentile society (societos) as distinguished from a political society
or state (rivit«). The difference between the two is wide and fundamental.
There was neither a political society, nor a citizen, nor a state, nor any
civilisation, in America when it was discovered. One eiitire ethnical period
intervened between the highest American Indian tribes and the beginning
of civilisation as that term is commonly understood. 1
s
There were six Iroquois tribes, speaking six dialects. Four of
them were each divided into two phratries and eight clans.
The other two had no phratries and only three clans. 2 Their
common origin is shown by the clan names, three of which
occur in all six tribes, while only two are confined to a single
tribe.
All the clans, with one exception, are named after animals.
These are the clan totems. It is told, for example, that one
hot summer day, after the pool in which it lived had been
1 Morgan AS 65. The Iroquois have been reinvestigated by Quain,
who holds that their highly developed military organisation was promoted
by contact with European colonists (245-7).
2 Morgan AS 6 g.
Ill
88 STUDIES IN ANCIENT GREEK SOCIETY
dried up by the sun, a turtle threw off its shell and grew into
a man, the ancestor of the clan that bore the turtle’s name and
emblem . 3
Table IV
THE LEAGUE OF THE IROQUOIS
Tribe Phratry Clans
Seneca
I
Bear
Wolf
Beaver
Turtle
n
Deer
Snipe
Heron
Hawk
Cayuga
I
Bear
Wolf
Turtle
Snipe
Eel
n
Deer
Hawk
Beaver
Onondaga
I
Wolf
Beaver
Turtle
Snipe
Ball
n
Deer
Bear
Eel
Tuscarora
I
Bear
Beaver
Great Turtle
Eel
n
Grey Wolf YellowWolf Little Turtle Snipe
Mohawk
Bear
Wolf
Turtle
Oneida
Bear
Wolf
Turtle
In Morgan’s time the exogamous unit Was the clan, but
tradition said it had once been the phratry, and this is con- .
firmed by the Iroquois word for ‘phratry’, which means a
‘brotherhood’. Clans of the same phratry were ‘brother’ clans; -
clans of different phratries were ‘cousin’ clans . 4 The Senecas
asserted that in the beginning their tribe had only two clans,
Bear and Deer, which later divided, the original units sur-
viving as the senior clans in their respective phratries.
The clan had a common residence, the long house’, sur-
rounded by gardens, and over the entrance was carved a device
representing the clan totem . 6 The house and gardens were
managed by the women, while the men occupied themselves '
with hunting and fighting. Tillage was done with the hoe, .
and the staple crop was maize. After an interval ranging from
ten to twenty years the soil became exhausted, and the tribe
moved to a new settlement . 6
3 E. A. Smith 77.
4 Morgan AS 90.
6 Morgan LI 318, where, writing before his discovery of the gens,
he calls it a ‘tribal device.’
6 Hale 50, Frazer TE 3. 3-4.
Ill ' FROM TRIBE TO STATE 89
Descent and succession were matrilineal. Each clan had its
own set of personal names, any of which might be bestowed
on a child provided it was not borne by a living member of the
clan . 7 A man’s personal effects .were distributed among his
maternal uncles, brothers, and sisters’ sons. They could not be
inherited by his own children. A woman’s heirs were her own
children, her sisters, and her sisters’ children. By this means
the property of the clan was retained within the clan. The dead
were mourned by their own clanspeople, but the preparation
of the grave and the actual interment were carried out by other
clans. A person of note might be mourned by his whole
phratry, and in that case the funeral would be performed by
the other phratry. In Morgan's day the dead were buried in-
discriminately, but from various indications he inferred that
each clan had once possessed its own cemetery.®
The Iroquois observed six annual festivals, which were
superintended by a prescribed number of officiants, male and
female, elected from each clan. They had no distinctive clan
cults, their place being taken by the ritual of secret societies
formed on die clan model. This is a general characteristic of
the Amerindian tribes, though in some the totemic increase
ceremony can be recognised in a modified form. The buffalo
dance of the Mandans, for instance, performed seasonally for
the propagation of that animal, differs from type only in not
being the perquisite of a particular clan.®
The clan had the right to adopt strangers, who were thereby
admitted to full membership as ‘brothers' or ‘sisters’ of the
persons responsible for their adoption, and received a clan
name. Captives were either adopted or put to death. Slavery
was unknown. 1 ®
The clan was responsible for the conduct of its members
and for protecting their interests. In the event of one of its
number being killed by a member of another clan, it lodged
against that clan a formal complaint and a demand for satis-
faction. If acceptable compensation were offered — usually a
payment in kind — the affair was at an end. If not, an avenging
party was appointed to pursue the manslayer and kill him. If
7 Morgan AS 77-80. 8 lb. 74-5, 83-4, 96, cf. Frazer TE 1. 75.
® Frazer TE 3 .137, 472. 18 Morgan AS 80-1 .
Ill
90 STUDIES IN ANCIENT GREEK SOCIETY
the two parties belonged to different phratries, the suit would
be taken up by the phratry on behalf of the clan concerned . 11
There was no recognised procedure for homicide within the
clan, and offences of this kind were extremely rare. In the
absence of private property the main incentive to such crimes
was wanting, and a positive deterrent was provided by the
spirit of intense solidarity that animated the clan.
The clan had its own chief (sachem) elected by the free vote
of the adults of both sexes. He was appointed for life, but
might always be deposed if he failed to satisfy the electors . 12
' The office tended to be hereditary, passing at the holder's
death to one of his brothers or a sister's son, and among die
Iroquois it was confined to men, but it is doubtful whether
this restriction was very ancient. The Winnebagoes of Wis-
consin observed the rule that, failing a brother or a sister’s son,
the succession passed to the nearest female relative on the
mother’s side . 13
Each tribe had its own territory and its own tribal council,
which met in public to decide questions of war and peace and
to ratify the election of clan chiefs, on which it had a veto.
Its decisions had to be unanimous. It was composed of the
clan chiefs, together with a number of war chiefs elected for
personal bravery, and also a special category of chiefs, whose
office was hereditary in particular clans and whose function
was to represent the tribe on the council of the confederacy . 14
This last body, the supreme organ of the Iroquois, was
composed of the special chiefs just mentioned. It too met in
public and was subject to the rule of unanimity. The consent
of all six tribes was required before it could act . 16 The actual
conduct of military operations was entrusted to two supreme
war chiefs, elected from the Wolf and Turtle clans of the
Senecas . 1 6
Morgan has some instructive remarks on the manner in
which these tribes had separated from the parent stock and
subsequently reunited:
New tribes and new gentes were constantly forming by natural growth;
■and the process was sensibly accelerated by the great expanse of the American
11 lb. 77, 95. 12 Ih , 70-3. 13 161-2. 14 lb. 113-20.
16 #.135. 16 lb. 1 50— 1.
Ill
FROM TRIBE TO STATE
9 *
continent. The method was simple. In the first place there would occur a
gradual outflow of people from some over-stocked geographical centre,
which possessed superior advantages in the means of subsistence. Continued
from year to year, a considerable population would thus be developed at
a distance .from the original seat of the tribe. ... A new tribe was thus
created. ...
When increased numbers pressed on the means of subsistence, the surplus
removed to a new seat, where they established themselves with facility,
because the government was perfect in every gens and in any number of
gentes united in a band. . . .
The conditions under which confederacies spring into being and the
principles on which they are formed are remarkably simple. They grew
naturally with time out of pre-existing elements. Where one tribe had
divided into several and these subdivisions occupied independent but con-
tiguous territories, the confederacy reintegrated them in a higher organisa-
tion on the basis of the common gentes they possessed and of the affiliated
dialects they spoke. The sentiment of kin embodied in the gens, the common
lineage of the gentes, and their dialects still mutually intelligible, yielded
the material elements for a confederation. The confederacy, therefore, had
the gentes for its basis and centre and the stock language for its circum-
ference. 17
We see how perfectly the tribal system was adapted to a society
constantly on the move. The multiplication of tribes was
simply a continuance of the process of self-division that had
created the tribe itself. But in the confederacy this movement
is reversed, and it is at this point that we observe, in the office
of the supreme war chiefs, the first departure from the principle
of equality. In the Iroquois League the tribes are about to
merge in the higher but class-divided unit of the state.
The League was designed for war. It was formed in New
York State after the expulsion of the Algonkins . 18 The
Iroquois had then reached the limit of free expansion at the
existing level of production. But, being still at the stage of
migratory agriculture, they only fought for land. If, before the
formation of the League, their agriculture had been more
advanced, they would have become sedentary, like the Village
Indians of Central America; or alternatively, if they had been
able to develop their agriculture under the League, they would
doubtless have used that instrument for subjecting other
tribes to some form of exploitation, as was done by the Aztec
17 lb. 105, 125. 18 lb. 169, SCA 150-1.
Ill
92 STUDIES IN ANCIENT GREEK SOCIETY
League in the more confined area of Mexico. As it was, their
progress was cut short at .this point by the followers of
Columbus.
2. The Roman Tribal System
Every Roman, at least every noble Roman, had three
names — a nomen or 'name', a prcenomen or forename', and a
cognomen or 'surname'. The prcenomen was personal; the nomen
denoted his gens or clan, the cognomen his familia or family.
Gaius lulius Czesar belonged to the Caesar family of the Gens
Iulia.
The familia was a subdivision of the gens. It comprised the
paterfamilias, his wife, his sons and unmarried daughters, his
sons' sons and unmarried daughters, his slaves, and other
household property . 1 0 The gens was a group of familice descended
in the male line from a common ancestor. The word familia
denoted originally property in slaves (famuli)— that is, acquired
goods as distinct from the collective property of the gens.
The property of an intestate passed in the first instance to
his wife and children; in default of children, to his direct
descendants in the male line; then to his agnatic kindred, con-
sisting of his brothers and unmarried sisters and his father’s
brothers and unmarried sisters; and finally, failing all these,
to his^ns. If we reverse these rules of priority, we have them in
their historical order, marking successive encroachments on
the common ownership of the gens. The married sisters and
daughters were excluded because the wife became by marriage
a member of her husband's gens.
The early history of Roman marriage is obscure, and - any
reconstruction is only tentative. Under the early Republic
there had been three forms of matrimony — usus, confarreatio , and
coemption The first was mere cohabitation. It required no cere-
mony, was dissoluble at will, and made no provision for the
transmission of property. It resembled the loose matriarchal
unions of the early Etruscans, to be described presently , 21 and
belongs probably to the time when plebeian marriages and
19 Morgan AS 293-5, Joiowicz 122.
80 Westrup RFA; Joiowicz 113-6, 243-4.
21 Sec below p. 142. On the antiquity of usus see Westrup 34-79*
Ill
FROM TRIBE- TO STATE
93
plebeian property rights had not been recognised by the
patricians. Tne patrician form was confarreatio, a deed of
transfer placing the bride under her husband's authority.
Coemptio was the corresponding plebeian form — a deed of
purchase giving the husband a contractual right to the pos-
session of his wife. Later, when the distinction between
patricians and plebs had disappeared, these forms were super-
seded by a union as loose as the ancient usus, but by this time
the interests of private property were secured by the right of
testamentary disposition.
The intention behind these patriarchal patrician unions is
quite clear:
If, says Cato, thou dost take thy wife in adultery, thou mayest kill her
without trial and with impunity; but, if thou dost commit adulteiy thyself,
she shall not and dare not so much as lay a finger upon thee. 22
Confarreatio circumscribed the woman's liberty in order to
safeguard the succession from father to son, and coemptio ex-
tended the same principle to the lower orders. They show how
formal matrimony was brought into being by the growth of
property as a juridical limitation of the old tribal rights:
The first class opposition that appears in history coincides with the de-
velopment of the antagonism" between man and woman in monogamous
marriage, and the first dass oppression coincides with that of the female sex
by the male. 28
Since the familia was a subdivision of the gens, the family
name was called the cognomen, a 'surname' or additional name.
The nomen, without qualification, denoted the gens. Again,
■while familia is a late word, connoting acquired property, gens
and nomen, 'kin' and 'knSW*, derive, as we have seen, from
the primitive clan, in which the kinsman had been known by
his clan name and clan emblem (p. 46). And when we look
into these nomina, their origin leaps to the eye. The Gens
Aquilia is the Eagle clan, Asinia is the Ass, Aurelia the Gold,
Caecilia the Lizard, Caninia the Dog, Capraria the Goat,
Cornelia the Cornel-tree, Fabia the Bean, Ovidia the Sheep,
Porcia the Pig, Valeria the Black Eagle, Vitellia the Calf etc.
22 Gell. 10. 23.
28 Engels UFPS 69.
Ill
94 STUDIES IN ANCIENT GREEK SOCIETY
There was more than idle fancy in the legend that Romulus
and Remus had been fed by a woodpecker and suckled by a
wolf. These animals are known from the evidence of tribal
and territorial names to have been sacred.
Here and there we meet totemic survivals in a more concrete
form. The Gens Quintia had a taboo on wearing gold orna-
ments; the Serrani family of the Gens Acilia forbade their
women to wear linen . 24 The Torquati of the Gens Manlia wore
a distinctive necklace, the Cincinnati of the Gens Quintilia a
distinctive coiffure . 25 Similar customs abound among primitive
in circumstances that place their totemic origin beyond
gens had its own chief ( princeps ), 28 its own shrine (sacel-
lum), its own cemetery, 2 ? and in early times its own land . 28
When the Gens Claudia migrated to Rome from the Sabine
country they were allotted a burial ground near the Capitol
and an estate on the banks of the Anio . 20 The gentile cult was
addressed to the genius, the ancestral spirit as such, or to one of
the public deities distinguished by the gentile eponym —
Silvanus Naevianus of the Naevii, Diana of the Calpurnii, Veiovis
of the Iulii, etc . 30 The conversion of ancestral spirit into
eponymous deity marks the transformation of totem into god.
There is no record of a personal name associated exclusively
with a particular gens, but the story of Marcus Manlius, who
brought such disgrace on the Manlii that they banned the
name Marcus, shows that the gens had a say in the naming of its
members . 31 Its consent was also required for the adoption of a
son, who thereby assumed his adoptive father’s nomen and
cognomen. The ceremony of adoption is described as an imitation
24 Plin. NH. 33. 21, 19. 8, cf. Frazer TE 2. 270, 295, 4. 24.
25 Suet. Cal. 35.
28 Cic. Fam. 9. 21. 2, Fest. 61, D.H. AR. 6. 69. 1.
2 7 Cic. Leg. 2. 22. 55, Off. 1. 17. 55, TD. 1. 7, Arch. g. 22, Val. Max. 9.
2. 1. Suet. Ncr. 50, Plu. Topi. 23, D.C. 44. 51, Veil. 2. 1x9. 5, CIL. I.
65-72. 375 -
28 T. Mommsen 1. 39, 74.
29 Suet. Tib. 1.
30 CIL. 6. 645, Cic. ic har. 32, CIL. 1. 807. The gins observed its own
feast-days: Macr. Sat 1. 16. 7, D.H. AR. 2. 21, 9. 19, Fest. 315.
31 Liv. 6. 20. 14; Darcmberg-Saglio 2. 2. 1510.
peoples
dispute.
Each j
Ill
FROM TRIBE TO STATE
95
of childbirth. 82 This idea is virtually universal. Adoption is
simply a special rite of initiation. 88 The stranger dies as a
stranger and is born again as a member of the clan.
The solidarity of the gens appears in a story of the Fab ii,
who, over 300 strong, fought a war against Veil all on their
own. 84 When Appius Claudius was thrown into jail, all the
Claudii went into mourning, including one who was his
personal enemy. 88 The gens was also expected to assist any of
its number who fell into poverty or distress. 88 The connection
between gentilis and generosus, ‘kinship’ and ‘kindness’, is
common to many languages and of all clan ties is the most
persistent. I have heard of Irishmen stranded abroad appealing
to complete strangers on the strength of a common surname,
just as Hardy's Tess visited the D’Urbervilles to ‘claim kin’.
That the gens was exogamous is nowhere expressly stated, but
we know that the Romans disapproved of the marriage of near
kin; 87 and if the rule had been observed from time" immemor-
ial, that in itself explains why, as a customary law, it was
never written'down.
We are told that in early days there had been 300 gentes
divided equally into thirty curia. 3 * The curia, which Greek
writers always rendered as phratrla, 3 * is the phratry or group of
related gentes. Each curia had its own shrine under a priest
called the curio. The thirty curiones constituted a sacred college
under the curio maximus, elected by the comitia curiata . 4 ° This
waP the assembly of all the men capable of bearing arms — a
32 Carp. Gloss. Lat. 4. 304. 44, Plin. Pan. 8 . 1, cf. Cassiod. Var. Ep. 4. 2.
88 E. S. Hartland in Hastings 1. 106, Gronbech 1. 305, Kovalev-
sly 125, Russell 2. 237, cf. John 3. 4-5, Rom. 8. 12-7. The Christian
baptism is at once a regeneration (p. 46 n. 38) and an adoption: ‘It hath
pleased thee to regenerate this infant and to receive him for thine own child
by adoption.* See further Eisler OF 63-5, Frazer FOT 2. 27-38.
84 Liv. 2. 48-50.
88 Liv. 6. 20. 2-3.
88 Liv. 5. 32. 8-9, D.H. AR. 2. 10. 2.
87 Plu. M. z 6 ^d, z 8 yd. On the problem of enuptio gentis, which does
not really bear on the present question, see Engels UFPS 138-41, Kagarov
FEPRR 637-40.
88 Liv. x. 13. 6, Plu. Rom. 20.
89 D.H. AR. 2. 7. 3, 6. 89. 1, Plu. Rom. 20, Popl. 7, D.C, 1-34. 5-9.
49 Liv. 27. 8. 1.
Ill
96 STUDIES IN ANCIENT GREEK SOCIETY
real gathering of the clans*. It was so called because the voting
was by curia, each curia having one vote, decided by a majority
of its getttes. 4 * It determined all questions affecting the en-
franchisement of strangers and the transfer' of citizens by
adoption from one family to another . 42
Just as ten gentes formed a curia, so ten curie formed a tribus.
There were three tribes — the Ramnes, mainly of Latin stock;
the Taties, mainly Sabme; and the Luceres, which included an
Etruscan element . 42 Each had its own tribal chief and together
they constituted the tribal league known as the Populus
Romanus . 44
The supreme organ of the league was the senatus or council
of elders. The number of senators had been raised in very early
times, and Niebuhr conjectured that originally they were the
clan chiefs (principes gentium ). 4B The executive power was
vested in a rex or king, appointed jointly by the senate and
cmitia curiata . 4<s The rex was commander-in-chief, high
priest, and supreme judge. After the fall of the monarchy his
political functions were transferred to the newly created
consuls, but the royal priesthood survived in the office of rex
sacrorum .«
To all this modern historians adopt an attitude of unreasoned
scepticism. Thus, according to Jolowicz, ‘it is more than
doubtful* whether the cmitia curiata carried proposals laid
before it by the king, ‘nor do modern authorities' believe that
the Roman historians were right in thinking that the king *was
elected by the cmitia’, while ‘the idea of representation is so
alien to what we know of the composition of the senate in
historical times that it cannot be believed to have operated
even in the earliest period ’. 48 The Roman historians must
have been at least equally conscious of the contradiction, yet
they accepted the tradition, presumably because it was then
too strong to be denied. The word rex exists in a cognate form
and with the same meaning in the Celtic languages, and the
Celtic kingship was elective . 48 So too in' all probability were
14. 3,4.20.2. 42 Gai. 1. 99; Jolowicz 86, 119, 12.5.
43 Liv. 1. 13. 8. ^D.H. AH. 2. 47. 48 Niebuhr 1. 338-9.
40 Liv. 1. 17, 32. 1, 35. 6, Cic. Hp. 2. 12. 3. 47 Liv. 2. 2. 1, 6. 41. 9.
48 Jolowicz 16-7. 49 Hubert 220, Skene 3. 141.
Ill
FROM TRIBE TO STATE
97
the Gaulish councils, which the Romans themselves likened to
'their own senate, while some Gaulish tribes, which had in
addition to the council a distinct war chief, are analogous to the
Iroquois. 60 The trouble with this school of historians is that
they- are trying to explain the tribal institutions of early Rome'
without raising the question of what tribal society is.
3. Matrilineal Succession of the Roman Monarchy
The first king of the Populus Romanus was Romulus him-
self, the founder of the city. The main body of the Sabines
were then independent under their own king, Titus Ta tins.
Romulus was succeeded by a Sabine, Numa Pompilius, who
was a son-in-law of Titus Tatius. The next king was a Latin,
Tullus Hostilius, and he was succeeded by another Sabine,
Ancus Martius, a son of Numa’s daughter. Then came the
Etruscan conquest. Tarquinius Priscus, the next king, was an
Etruscan. His successor was Servius Tullius, a slave, either
Etruscan or Latin, who had married his daughter. From him
the succession passed to his son-in-law, Lucius Tarquinius, a
son of Priscus, and with him the monarchy ended.
In this tradition the royal office passes regularly in the female
line . 51 Ancus Martius is a son of his predecessor’s daughter,
Pompilia, implying that he succeeded through his mother.
Similarly Numa, the father of Pompilia, had married his
predecessor’s daughter; Servius Tullius married the daughter
of Priscus, and Lucius the daughter of Servius Tullius.
Romans of a later age are not likely to have invented a tradi-
tion so repugnant to their prejudices.
Succession from father-in-law to son-in-law is a recognised
mode of matrilineal inheritance. The office is held by males
but transmitted through females. The Iroquois rule, from
mother's brother to sister's son, rests on the same principle, the
difference being simply that the Roman presupposes a more
advanced development of matrimony. Now, if the kingship
passes from father-in-law to son-in-law, the queenship passes
60 Hubert 221-2.
si Frazer GB-MA 2. 270-2. The alternate succession of Latins and
Sabines may be compared with the Gaelic rule of tanistry: Skene 3. 150.
98 STUDIES IN ANCIENT GREEK SOCIETY III
from mother to daughter. Does this mean that the king reigns
in some sense on his wife’s behalf? We shall see in die next
chapter that it does.
The Etruscans are known to have been matriarchal. In some
of their epitaphs the name of the deceased is followed by his
Talk V
THE SABINE AND ETRUSCAN KINGS OF ROME
Titus Tatius
Tatia=Numa Pompilius
PompiIia= Martius
)
Ancus Martius
Tanaquil=Tarquinius Priscus
daughter=Servius Tullius
Lucius Tarquinius=Tul
father’s. This is patrilineal. In others both' parents are given.
This is ambiguous. In others only the mother’s name is
added . 62 Genus huic materna superbum nobilitas dabat, incertum de
patreferebat . 63 These epitaphs mark the decline of mother-right. .
Greek historians tell us that the Etruscans had ‘wives in
common’ and ‘their children did not know their own fathers ’. 64
They say exactly the same of the prehistoric Athenians . 66 It is
simply a conventional description of the matriarchate, in
which the woman is free to marry the man of her choice and as
many as she pleases. There is no question of adultery — that
was the man’s invention — and she retains control of her
62 R. S. Conway in CAH 4. 405. So in Lycian inscriptions: CIG. 4266b,
4316a, 4278, 4215, 4300. The Latin parens ‘parent' meant originally
‘mother’: Odgers LP.
63 Verg A. 11. 340-1.
64 Theop. 222, cf. Liv. 4. 2. 6.
66 See below p. 142.
Ill
FROM TRIBE TO STATE
99
children without regard to their paternity. So among the
Lycians of Anatolia, another matriarchal people, the child of a
freeman by a female slave was servile, but the child of a male
slave by a freewoman was free. 66 This explains how it was that
by marrying an Etruscan princess a slave became king of the
Eternal City, and how he was succeeded by her brother, who
had consolidated his position by marrying her daughter.
Among the Sabines memories of mother-right survived in
the stories of Drances, who became chief of the Rutuli through
his mother, and of Camilla, the warrior queen of the Volsci. 67
Sabini, Rutuli, and Volsci all belonged to the same stock. The
rape of the Sabine women is usually explained as a case of mar-
riage by capture,' and so it was, though it is recognised that this
mode of getting wives is less common than was at one time
supposed. But, if the Sabines were matriarchal, it is possible
that what the Romans were really after was not so much the
ladies themselves as their estates.
In the Latin kings the matrilineal rule does not appear.
Does this mean that the Latins were patriarchal? If so, they
were already one step ahead of the other Italic tribes — the
first on the road to world conquest.
One more question: how is the Sabine matriarchate to be
reconciled with the evidence that the Indo-European stock was
patriarchal at the time of its dispersal? The answer lies in
Italian prehistory, which has yet to be uncovered by the spade.
We must remember that, being determined by economic
forces, rules of inheritance are liable to change. Some authori-
ties would connect the Italic peoples with the terramara cul-
ture, 68 which, being based on tillage, was probably matriarchal.
And in any case these peoples developed under Etruscan influ-
ence, which must have affected their native institutions. This
is a process we shall meet again in Greek prehistory.
4. The Populus Romanus
The gens , curia, trilus, and populus are the Iroquois clan,
phratry, tribe, and league. The Populus Romanus and the
66 Hdt. 1 . 173 . 5. The same rule obtained in ancient China: Wittfogel 400.
67 Verg. A. 11. 68 T. E. Peet in CAH 2. 568-74.
Ill
IOO STUDIES IN ANCIENT GREEK SOCIETY
League of the Iroquois are structures of the same type. There is
only one important difference.
The Iroquois tribes evolved by natural expansion and
federated by simple aggregation without violence to their
internal structure. The Roman confederacy was created' artifi-
cially out of heterogeneous elements by an arbitrary act. Its
artificial origin cannot be argued from the symmetrical group-
ing of gentes and curia, which may be due simply^to the
partiality of oral tradition for round numbers, but it is proved
by the word tribus, which, since it means a ‘third’, presupposes
the federation of the three tribes. The confederacy was formed
for the express purpose of organising the new settlement at
Rome. It illustrates what would have been the next step in the
history of the Iroquois, had they survived to adopt a sedentary
life. Indeed, as Morgan showed, it closely resembles the con-
stitution adopted by the Aztec League when they founded the
city of Mexico. 60 The Populus Romanus marks the point at .
which the clansman is about to become a citizen and the
tribal system converted into a state.
Conversely, the Iroquois tribes illustrate what the 'Roman
had ceased to be. In still earlier times the various offshoots of -
the Sabellian stock had been expanding through the Italian
peninsula in the same way as , the Amerindians covered North
America. Year by year, according to tradition, the Sabellian
tribes had sent out a band of newly initiated young men and
women to seek a new home.
One swarm of these emigrants, who took the ox of their god Mars as
their badge and omen, struck southwards into the glens round Bovianum,
the ‘ox-town’, where they became known later as die Samnites; a second,
devoted to the wolf ( hirpus ) pushed further in the same direction, and appear
as the Hirpini; a third, led by the woodpecker (pints), pressed north-east-
wards towards that part of the Adriatic coast, south of Umbria, which
became known after them as Picenum; while a fourth, dedicated more ex-
pressly to their own god Mars, formed the warlike tribe of the Marsi, near
the Fucinc Lake, in the heart of the Sabellian highlands, oo
The Populus Romanus was constituted deliberately for
die purpose of organising the new setdement on the Tiber.
Those .who like to think of this act, fraught with such
so Morgan AS 191-220, cf. Bancroft 2. 226-7. 00 Myres HR 19.
Ill FROM TRIBE TO STATE IOI
consequences for the future, as the work of one man, may not be
altogether mistaken. The invitation issued by Romulus to all
and sundry to come and join him in the asylum on the Capi-
toline has a parallel in Greek history. Some time in the sixth
century B.c. the Greeks of Kyrene appealed to the home
country for settlers to join them in a repartition of the soil.
This was effected by a reconstruction of their tribal system, the
newcomers being incorporated along with the old colonists in
- a confederacy of three tribes with common lands assigned to
each, and die whole procedure was carried out under the
supervisionof a specially appointed arbitrator . 61 From this we
see that in Greece and Rome alike, at the times in question, the
tribal structure was becoming a merely formal entity — an
empty husk — with only a nominal basis in actual consanguinity.
In Rome, under the Republic, it disappeared, but the Greek
city-states never shook it off. As long as they lasted, they con-
tinued to organise their citizens in tribes — an unconscious
testimony to the dependence of the present on the past.
61 Hdt. 4. 159, 161.
IV
GREEK TRIBAL INSTITUTIONS
i. JEolians, Dorians , and Ionidns
The Greeks recognised three branches of their race, associated
with their three main dialects. The ^Eolians inhabited
Thessaly and Boeotia together with Aiolis on the opposite
coast of Anatolia. The Dorians covered the east and south of
the Peloponnese and extended overseas to the southern
Cyclades, Crete, Rhodes, and the Carian coast. The Ionians
occupied Attica, the central and northern AEgean, and part of
the Anatolian littoral, which was known after them as Ionia.
The Dorians were the latest comers, and their tribal tradi-
tions are the fullest. They entered southern Greece at the end
of the second millennium B.C. At that time they were a league
of three tribes: the Hylleis, descended from Hyllos, a son of
Herakles; the Dymanes, whose god was Apollo; and the
Pamphyloi, ‘men-of-aH-tribes’, who worshipped Demeter . 1
They came from the highlands of Doris in Central Greece . 2
Doris lay between the mountain masses of Parnassos and Oita
at the head of die Kephisos, which flows down into die rich
Boeotian plain. To the south of Parnassos lies Delphi, the
great seat of Apollo, whose cult was brought there in pre-
historic times from Crete and S.W. Anatolia . 3 Oita was t;he
scene of the death of Herakles , 4 * the hero of Boeotian Thebes.
There were prehistoric cult centres of Demeter at Lebadeia in
die Kephisos valley and at Pyrasos in southern Thessaly, which
the Dorians are said to have occupied before moving south . 6
The name of the third tribe and the three tribal cults suggest
that die Dorian League was an artificial construct, like the
1 Paton 341, Meillet AHLG 96.
2 Str. 475-6, cf. 383, Hdt. 1. 56, Paus. 5. x. 2.
3 See below pp. 293-4.
4 Apld. 2. 7. 7.
6 Paus. 9. 39. 1-5, II. 2. 695-6, Hdt. 1. 56.
IV
GREEK TRIBAL INSTITUTIONS
io 3
Table VI
PREHISTORIC GREEK CHRONOLOGY
Date
Main land
Cyclades
Crete
Egypt
3300
Neolithic
Early Helladic
Neolithic
3200
Dynasties
i-m
3100
3000
2900
Early Cycladic
Sub-neolithic
2800
Early Minoan I
IV
2700
Early Minoan II
V-VT
2600
2500
2400
Early Minoan III
VH-X
2300
2200
2100
Middle
Cycladic
Middle Minoan I
XI
2000
1900
Middle Helladic
Middle Minoan II
XII
1800
1700
Middle
Minoan HI
xrn-xvn
1600
1500
Late Helladic I
Late Cycladic
Late Minoan I
1400
Late Helladic II
Late Minoan II
xvni
1300
Late Helladic III
Late Minoan III
XIX
1200
1100
Sub-Mycenean
XX
See Pendlebuiy 301. The Late Helladic periods are also known as the
Mycenean.
Populus Romanus, formed in Central Greece under the in-
fluence of the prehistoric cultures of Delphi and Bceotia.
When they settled in the Peloponnese and overseas in the
southern .Egean, they took their tribal organisation with
them . 8 This need not mean that all three tribes actually par-
ticipated in each movement. It is more likely that the system
8 Hie three tribes are recorded in most Dorian settlements, but it
appears they were not established at Telos (ZG. iz. 3. 38). We hear of one
of them, the Dymanes, migrating by itself from Troizen to Halikamassos
(St.B, *MiKapvaaa6s).
IV
104 STUDIES IN ANCIENT GREEK SOCIETY
was thrown into confusion by the migrations and reconstituted
in their new homes on the traditional pattern.
The Ionians had four tribes. The names — Aigikoreis,
Hopletes, Argadeis, Geleontes — have not been explained.
This is not surprising, because there are many instances in
modern ethnology of tribal names being acquired quite for-
tuitously. 7 All we know of their cults is that the Geleontes
worshipped Zeus Geleon, and that the patron of the League'
was Poseidon Helikonios, the god of Mount Helikon in-
western Boeotia. 8 The League was certainly older than the
colonisation of Ionia, because the same four tribes are found in
Attica. When and how it came into being is a problem to
which we shall have to return later. 8
2 . The Attic Trihal System
The Greek words for tribe, phratry, and clan are, in Attic,
phyU, phratria, and ge'nos. The phyle' is properly a ‘growth’ or
‘stock’. The phratria, like the Iroquois term for the same unit
(p. 88), is a ‘brotherhood’, implying a collateral relationship
between its constituent clans. The genos, corresponding to the
Latin gens, goes back to a root deeply imbedded in the Indo-
European languages.
In Aaolic and Doric ginos is replaced by patra , ‘fatherhood’,
implying descent in the male line. 10 In Attic, besides gennetes,
the regular word for ‘clansman’, we find homogalaktes, ‘ fed on the
same milk’, implying descent in the female line. 11 These are
the sort of variations we should expect if there had been
changes in the mode of succession.
As the tribal system decayed, these words came to be used
loosely with wider applications. We find phyle (phylon) applied
generally to any consanguineous stock, sometimes apparently
7 Morgan AS x 14.
8 IG. 28. 1072, Hdt. 1. 148.
8 See p. 392.
10 ^ P* 7* 5» 8. 38, N. 4. 77, 6. 36, 8. 46, 11. 20, 1. 6. 63, Hdt. 2
143. i, SIG . 438.
11 Hsch. Suid, 6py£wvccs, Arist, Pol . 1252b. 6, Poll. 6. 1 56.
IV GREEK TRIBAL INSTITUTIONS I05
even to a clan . 12 Qinos was still more unstable. It came to mean
‘kinship’, ‘kind’, ‘birth’, ‘breed’, ‘race* without reference to its
tribal origin . 13 The same thing of course has happened in
modern languages. The old words for these units have been
lost, and the new ones adopted by ethnologists, such as ‘tribe*
and ‘dan*, are often used very vaguely . 14 But the ancient Greeks
stood much nearer to tribal society than we do, and though
they sometimes used the words loosely they never confused
the things themselves.
Aristotle says that the early Athenians were organised in
four tribes, each tribe containing three phratries, each phratry
thirty clans, and each clan thirty men. He adds that the four
tribes corresponded to the seasons, the twelve phratries to the
months, and the thirty clans in each phratry to the days of the
month . 15 That there were three phratries in each tribe is per-
fectly credible, and the distribution of the clans is not more
schematic than the Roman, but what is the meaning of the
parallel with the calendar?
Under the democracy the number of tribes was raised to ten,
and the civil year was divided into ten periods, during each of
which a standing committee elected from one of the tribes was
in session. If this principle of tribal rotation was a new one, we
may suppose that in the tradition recorded by Aristotle it has
been projected retrospectively into the past. But it may be
doubted whether it was new. The democratic constitution was
designed to reproduce the external features of the old system
which it had superseded . 18 If the four old tribes had functioned
separately for certain purposes in successive quarters of the
year, such an arrangement would have been wholly in keeping
.with the ritual co-operation characteristic of tribal society
everywhere. In that case the only unhistorical element is the
12 Od. 14, 68, Hdt. 4. 149. 1.
13 Hence PI. Phlb. 30c! sch.: 'yewfiToti are not individuals related by blood
or birth but members of the ytvri grouped in phratries'; Harp. ytwfjTcn:
'the term yswiiTai, members of the same y£vos, was not applied to kinsmen
in the simple sense, kinsmen by blood (delete ol before {§ oThotos), but to those
distributed in the so-called y£vT|\ cf. Poll. 3. 9.
14 Cf. Morgan AS 64.
16 Arist. fr. 385.
18 G. Thomson AA 207-8.
IV
Io6 STUDIES IN ANCIENT GREEK SOCIETY
extension of the parallel from die phratries to the clans— a
formal simplification to which oral traditions are always
liable. »
There remains the total of thirty men to a clan, which is not
explained by the calendar. The figure is probably based on a
conventional estimate of man-power calculated for conscrip-
tion or taxation, like the Anglo-Saxon ‘hundred’, which re-
presented nominally a hundred heads of households. This
analogy was cited by Grote. Another, not available in his day,
is furnished by inscriptions from Samos, where each tribe was
divided into ‘thousands’ and each ‘thousand’ into ‘hundreds’. 1 ’
Whatever interpretation may be placed on the symmetry of
this system and the parallel with the calendar, the kernel of
the tradition, concerning the organic relation between the three
units, is unaffected. Here Aristotle is at one with Polybius,
Dionysius, Plutarch, and Dio Cassius, all of whom treat phyle ,
phratria , and genos as equivalent to the Latin tribus, curia , and
gens.™ The tribe was a group of phratries, the phratry a group of
clans. On this point the ancient authorities are unanimous,
and, since the same result has been reached by modern re-
search on the tribal system in all parts of the world, we may
say that of all the facts relating to the social organisation of
prehistoric Greece there is none more firmly established.
It is against this solid background that we must set the view of
recent historians, who, ignoring the external evidence, have
been at pains to refute the testimony of Aristotle. According
to E. A. Gardner, writing in the Cambridge Ancient History,
the early Athenian tribes consisted of ‘so many independent
war bands’; the phratries, which in origin ‘appear to have
been voluntary associations, composed in the first instance of
comrades-in-war*, were admittedly subdivisions of the tribes;
but the dans, described as ‘sectional assodations’ constituting
‘artificial aggregates of families rather than one interrelated
group’, were not subdivisions of the phratries . 19 The quality
17 Grote 3. 54, Vinogradoff GM 144; Supp. Epig. Gr. 1. 350, 354-51
362 etc. Cf. Exod. 18. 2i— 2, Thompson 49. Aristode's 'thirty' is probably
connected with the -rpiaxis, a sub-division of the Snuos: SIG. 9x2. 19.
18 See p. 95 n. 39.
19 E. A. Gardner in CAH 3. 584-5, cf. F. E. Adcock in CAH 3. 688.
IV GREEK TRIBAL INSTITUTIONS io?
of the arguments invoked in support of these statements is
worthy of attention.
In Athens, as in many odier Greek states, probably in all,
the tribes functioned as units of the army . 20 In the Iliad the
Achseans are marshalled ‘tribe by tribe and phratry by phra-
try ’. 21 The military functions of the tribal system are of
course as old as warfare, but the system is older still. The idea
that it had its origin in warfare is a gratuitous invention.
Regarding the relation between the phratry and the clan,
Gardner affirmed that ‘a decisive argument against Aristotle is
supplied by an extant fragment of an early Attic law, which
prescribed that the phratries must admit not only members of
the clans but other categories of citizens as members ’. 22 This
law belongs to the sixth century, when the old Attic system
was breaking down. So far from proving that Aristotle was
wrong in describing the phratry as a group of clans, it proves
that he was right, because, if non-clansmen had not been
previously excluded, there would have been no need to pass a
law enforcing their admission. Laws are not made to compel
people to do what they have always done of their own volition.
It might make things easier for the historian if they were, but
they are not.
‘Again’, the argument continues, ‘it is certain that the
clans were not subdivisions of the phratries. As a general rule,
the members of each clan did not all belong to the same
phratry but were distributed at random among these groups,
and the case of the Eteoboutadai, a clan whose members were
included en bloc in one and the same phratry , 23 must be re-
garded as exceptional. It follows that the clans stood in no
definite relation to the phratries.’ So far as it goes, this state-
ment is perfectly correct, but, since it purports to describe the
state of affairs before the democratic revolution, when the old
system was still in being, the unsuspecting reader should have
been warned that the evidence on which it rests is taken from
the period after the revolution, when the old system had been
abolished; and once this small but necessary adjustment has
been made, the correct conclusion is seen to be the opposite of
20 Is. 2. 42, Hdt. 6. hi. i, Th. 6. 98. 4. 21 II. 2. 362-3.
22 See below p. 112. 23 Aischin. 2. 147.
IV
Io8 STUDIES IN ANCIENT GREEK SOCIETY
that which is here described as certain. We have just observed
that membership of the phratry, which carried civic rights,
was thrown open in the sixth century to non-clansmen. That
was the first blow to the phratry. The second came at the end of
the century. Then, under the new democratic constitution, ,
civic rights ceased to depend on the phratry at all. The result
was that both phratry and clan, divorced from political life, fell
into decay. And in these conditions the circumstance that
commands attention is not the ‘general rule’-r-the severance
of the organic link between these units — but the exception to
it, which is here dismissed as fortuitous. Of all the Athenian
clans the Eteoboutadai, or Boutadai, were the most old-
fashioned and exclusive. They boasted of having the blood of
Erichthonios the earthborn running in their veins ; 24 their
hereditary privileges included, among other ancient priest-
hoods, the cult of Athena Polias, the patron goddess of die
state ; 26 in the sixth century they had rallied the other big
landlords behind the banner of conservatism in opposing the
reforms demanded by the merchant class ; 28 and at die time of
the democratic revolution they seem to have still retained some
at least of their ancestral estates, because one branch was then
still resident at Boutadai , 27 which, as its name shows, was
their original seat. Accordingly when, a century later, we find
the whole of this true-blue, die-hard clan enrolled in the same
phratry, the proper inference is that it was proudly adhering to
what had once been the general rule.
Lasdy, we are assured that ‘the artificial character of the
clans is expressly attested by ancient writers; it is also indicated
by the obviously mythical character of the ancestors from whom
they drew their name, and by the longevity of several clans
which maintained an unbroken existence to the days of the
Roman Empire’. The only sense in which ancient writers
bear witness to die artificiality of the clans is that admission
could be obtained by adoption; but this is true of all dans the
world over, no distinction being drawn in p rimi tive thought
24 Apld. 3. 14. 8, Plu. M. 843c.
26 Apld. 3. 15. 1, Paus. 1, 26. 5, Aeschin. 2. 147.
28 Hdt. 1. 59-60.
27 Plu. M. 841b.
IV
GREEK TRIBAL INSTITUTIONS
IO9
between birth and re-birth. Similarly, if the common descent
of the Greek clan is disproved by the fact that the eponymous
ancestor is usually mythical, the common descent claimed by
totemic clans at the present day, and confirmed in many cases
by extant genealogies, must also be a myth, because in these
the ancestor is usually an animal or vegetable. As for Gardner’s
parting shot, that their artificiality is indicated by their
longevity, one can only reply that in his ‘home of lost causes’
there are still gentlemen who 'came over with the Conqueror’.
This attitude to the problem was pardonable in Grote, who,
writing before Morgan, had some reason for concluding that
‘the gentile and phratric unions are matters into the beginning
of which we cannot pretend to penetrate ’, 28 but Morgan's
discoveries, not to mention other achievements of social an-
thropology, have been available for half a century; and con-
sequently, when we find that the effect of dissolving Aristotle's
clear delineation of the Attic tribal system into the independent
war bands, voluntary associations, and artificial aggregates of
the Cambridge Ancient History is to obscure what had been
elucidated, we cannot help wondering why Grote’s successors
should so resolutely prefer darkness to daylight. Can it be that
this dusty little cupboard, into which he could not, and they
will not, pretend to penetrate, contains a skeleton — the origin
of the family, private property, and the state?
3. The Household
An Athenian citizen was known officially by his personal
name followed by his father's and that of his deme (demos).
The deme was the urban or rural district in which he had been
registered at birth. In other states we find the clan name in
place of the patronymic . 28 There was no cognomen to denote the
fainily. The Greek equivalent of the Jamilia was the otkos,
■ ‘household’, or anchisteia, denoting the ‘next-of-kin’ within the
wider circle of the ge'nos . 30 It consisted of the founder and his
children,' his sons’ children, and the children of his sons’ sons.
When he died, his estate was inherited by his sons, who might
either hold it jointly or divide it, but in either case they owned
28 Grote 3. 58. 29 CIG. 3064. 30 H. E. Seebohm 54-64, 88-97.
1 IO STUDIES IN ANCIENT GREEK SOCIETY IV
it in common as co-heirs. If one of the sons had predeceased
the founder, his share went to his own sons, or, if they were
dead too, to his grandsons. In the fourth generation, however,
the estate was finally divided among the founder’s great-
grandsons, each of whom founded a new household . 31 This
limitation applied to the institution in all its aspects. The
duty of maintaining the founder in his old age and tending his
grave devolved on the sons, grandsons, and great-grandsons . 32
The responsibility for prosecution in cases of homicide ex-
tended as far as the children of the victim’s first cousins, who,
as descendants of the same great-grandfather, were the most
remote relatives comprehended in the household . 83 If a man
died without issue, his heirs were, in the order stated, his
father, his brothers and their children, his father’s brothers’
,children, and the children of his father’s brothers’ sons. If
none of these survived, the estate passed, not to the remoter
descendants, but to his mother’s household . 34
-The divergences between the oikos and the famili a arise from
the immaturity of the Attic law of property as compared with
the Roman. The limitation to the fourth generation is an
archaic feature which the familia probably lost when the estate
became alienable. The right of free testamentary disposition
was not recognised in Attic law, and so the estate was at least
nominally inalienable . 86 Again, the Roman wife was a member
of her husband’s familia and consequently a co-heir to his
estate. The Athenian wife, on the other hand, remained in the
guardianship of her own oikos. Accordingly she had no share
at all in her husband's inheritance, the only exception being
that, if his oikos was extinct, the estate went to her own . 36
81 lb. 56-64.
32 Is. 4. 19, 8. 32, iEschin. 1. 13.
33 D. 43. 57, Pi. Leg. 871b, cf. 877c. The same limitation applied
to the admission of kinswomen to the house of the dead: D. 43. 62, 57. 66,
SIG. 12 1 8. In both cases the motive was to prevent a clan vendetta: see
p. 481.
34 Is. 7. 22, 11. 1-2, D. 43, 51. In D. 43. ii-z the plaintiff, a grandson
of the deceased s first cousin, is adopted by his grandfather and so brought
within the oIkoj. '
36 This right is unknown to primitive law: Diamond 248-50.
36 H. E. Secbohm 27-8.
IV GREEK TRIBAL INSTITUTIONS III
Most of our information about die oikos comes from Athens’,
but similar rules of inheritance, with the same limitation, are
found in' the laws of Gortyna (Crete), the only other code that
has survived, 87 and the principle of joint succession underlies
the Homeric myth of the division of the world among the
Sons of Kronos, Zeus received the sky, Poseidon the sea,
Hades the darkness; the earth and Olympus remained com-
mon. 38 The three elements represent the personal estate,
which is divided; the real estate — the land and die house — is
held joindy.
The origin of the oikos has been explained by H. E. Seebohm:
It was extremely improbable that a man would see further than his great-
grandchildren bom to him before his death. And it might also occur in time
' of war or invasion that his sons and grandsons might go out and serve as
soldiers, leaving the old man and his great-grandchildren at home. . . Thus,
especially in cases where the property was held undivided after the father's
death, we can easily see that second cousins (i.e. all who traced bade to the
common great-grandfather) might be looked upon as forming a natural
limit to the immediate descendants of any one oikos and as the furthest re-
moved who could daim shares of the ancestral inheritance. After the death
of the great-grandfather, the head of the house, his descendants would
probably wish to divide up the estate and start new houses of their own. The
eldest son was generally named after his father's father, and would carry on
the name of the eldest branch, and would be responsible for maintaining the
rites at the great-grandfather's tomb. . . . Thus seems naturally to spring up
an inner group of blood relations dosely drawn together by ties which only
indireedy reached other and outside members of the g(nos. 88
Similar” types of household, with in some cases the same
limitation, have been found among the Celts, Germans,
Slavs, and Hindus; 40 and there is one indication of common
-origin. We saw that among the Indo-European terms of re-
lationship there was one — the term for the husband's brother's
wife — which is not referable to the classificatory system (p. 83).
This may now be explained as an innovation of the patriarchal
household, which included under the same roof in each genera-
tion subsequent to the founder a group of women related to
one another only by their marriage to a group of brothers. It
follows that even at this early date the Indo-European dan
37 Lex Gort. 5. 10-21. 38 I/. 15. 187-93. 89 H, E. Seebohm 54-5.
49 lb. 49-54, F. Seebohm EVC 351, Kovalevsky 60-100.
112 STUDIES IN ANCIENT GREEK SOCIETY ' IV
contained the germ of the modem family — a sign that the
dispersal of these peoples, as well as the collapse of their clas-
sificatory system, was due to the pressure of individual rights
of property.
4. Pre-Hellenic Clans in Attica
One of Gardner’s arguments against Aristotle was the early
Attic law prescribing the admission of non-clansmen to the
phratries (p. 107). The actual words are these: Tt shall be com-
pulsory for the phratry to admit orgeones and homogdlaktes’ . 41
Gardner did not attempt to identify these categories. He
simply took it for granted that they were non-clansmen. This
assumption cannot be accepted without further argument,
because Philochoros, who quotes the law, adds that the
orgeSnes and homogdlaktes ‘are what we call clansmen (gennetaCf .
The orgeonesrwexe members of a religious guild which met in
the deme to offer monthly sacrifices to the local god or hero. 42
These guilds were peculiar to Attica, and had an official status.
When a citizen adopted a son, he presented him to the fellow
members of his phratry (pkrdteres), the fellow members of his
deme (demit ai), and his fellow orgeones .* 3 The phratry was not a
territorial unit, but the demit ai and the orgeones both belonged
to the same locality. It may be conjectured that the orgeones
were the demitai acting in a religious capacity.
This evidence dates from after the democratic revolution,
when the demes were reorganised as units of local govern-
ment. They had of course existed before the revolution, but
simply as villages, without official status. We may infer that
the orgeones were a body of persons appointed by and from the
demitai to administer the village cult.
Nearly 200 Attic demes are known to us by name, and at least
thirty of them are clan names. 44 The deme Ifhilaidai, for ex-
ample, was situated near Brauron, where Phiiaios, eponym of
41 Philoch. 94, cf. Poll 3. 52.
42 Phot. 6pyESvss, Poll. 8. 107, AB. 1. 191. 27, 227. 15, SIG. Iioo.
24, 1101. 15.
43 Is. 2. 14* The dpyeCvEs are not always mentioned in this formula •
(Is. 7. 27), probably because they did not exist in all parts of Attica.
44 Pauly-Wissowa s.v. Sfjiioi.
IV GREEK TRIBAL INSTITUTIONS 113
the clan Philaidai, had landed in Attica . 46 In these instances at
least the deme was certainly a clan settlement. In addition we
find about twenty-five demes named after some species of tree
or plant — for example, Aigilia (wild oats), Hagnous (willow),
Marathon (fennel), Myrrhinous (myrtle), Rhamnous (buck-
thorn). These suggest local cults of herbal magic and tree-
worship, such as are known to have been widespread in pre-
historic Greece. The word orgein is probably connected with
Srgia, secret rites, ‘orgies’, and with orgds, a piece of consecrated
ground, tilled or untilled , 46 like the sacred grove of poplars that
stood outside the town of Ithaca . 47 Such groves still exist in
the Ionian Islands . 48 They are a regular feature of the village
in many parts of Europe and Asia, and in India they still
serve for the worship of the local earth goddess . 49
The earliest remembered inhabitants of Attica were Pelasgoi,
a non-Hellenic people to whom we shall be introduced later.
I suggest that the orgeones were originally the clansmen of the
Pelasgoi. These clans were matrilineal — hence the name
homogdlaktes, ‘fed on the same milk*. They lived in village
settlements, each with its sacred grove (orgds) for the main-
tenance of the clan cult (drgia). The Greek-speaking invaders
brought with them their -own tfibal system, from which these
aborigines were excluded. Hence, save in so far as they were
absorbed by the new clans, the old Pelasgian village cults sank
into obscurity. But they did not die out. In the sixth century
they formed a natural rallying point for vagrants, outcasts,
squatters, and other detribalised elements uprooted by the
appropriation of the land, and after the democratic revolution
they came into their own again, taking their place in the new
system of demes which superseded the old aristocratic clans.
46 PIu. Sol. 10.
48 Harp, dpytcovas. 'Opyec&v' stands to * 4 pyd as irerTpec&v to to 5 crpoc ( 7 G. 12.
I. 892).
47 Od. 17. 204-11, cf. 6. 291-4.
' 48 Ansted 191-5.
49 Baden-Powell 23, Russell 1. 44, Gurdon 33, Ehrenfels 96, cf. Earthy"
25, Rattray A 246.
H
IV
114 STUDIES IN ANCIENT GREEK SOCIETY
5 . Totemic Survivals: Snake-worship
The Athenian genos maintained an ancestral cult under the
direction of its drchon or chief 60 at a special shrine, and had its
own burial ground , 61 where the dead were worshipped as
heroes. The tendance of the dead, and probably the clan cult
as a whole, were based on monthly observances, like the village
6rgia. 62 Were these cults totemic?
The comparative study of totemism creates in favour of an
affirmative answer to this question a presumption so strong as
to place the burden of proof on the other side. Those who deny
the presence of totemic elements in Greek religion have only
been able to maintain their position by isolating the subject
from its proper context in the general history of religion. And
the result is that one of the most • conspicuous features of
Greek culture — the part played in myth and ritual by plants and
animals — is left unexplained.
We have seen how, by treating the clan as an ‘artificial aggre-
gate of families’, the Cambridge Ancient History envelopes the
origin of the family in the obscurity of an impenetrable past.
So in regard to the clan cult, avoiding the word totemism as
though it were indelicate, these authorities assert that ‘the
proximity of Greek religion to this hypothetical pre-deistic
stage of culture falls to the ground ’. 63 Apollon Lykeios is
admittedly a wolf-god , 64 but, if this wolf-god was ever a wolf,
it was so long ago that there is no need for him to poke his
nose into the picture. Even Nilsson, who has contributed so
much to Greek archeology, insists that ‘there is nothing in
Greek religion which necessarily demands a totemistic explana-
tion’ and that ‘it is unproved and doubtful whether totemism
ever existed among die forefathers of the Greeks ’. 66 It can be
shown in detail diat, when confronted with this problem,
Nilsson’s reasoning, usually so clear and cogent, breaks down.
60 IG. 2. 605, 3. 5, 97, 680, 702. This usage of fipxcov is not given in
Liddell and Scott.
si Plu. Them. 1, Hdt. 5. 61. 2, Paus. 1. 2. 4-5 (cf. Poll. 8. 103), IG. 2. 596;
D * 43 - 79 . 57 - 28.
62 S. El. 281, cf. Plu. M. 2g6£, SIG, 1218-0, A. Mommsen 3-5.
63 W. R. Halliday in CAH 2. 613.
64 lb. 2. 632. 66 Nilsson HGR 77-8, cf. Deubner 171. -
IV GREEK TRIBAL INSTITUTIONS 115
One of the cardinal characteristics of Greek religion, at-
tested continuously from Mycenean to Christian times, is snake-
worship. The subject of Greek totemism may be introduced
by a survey of these cults, which will also throw some further
light on totemism in general.
In Epeiros, always one of the most backward parts of the
country, there survived down to the Christian era a sacred
wood of Apollo. A number of snakes, believed to have sprung
from the dragon of Delphi, were tended there by a priestess,
who alone was permitted to enter the enclosure. She fed them
with honey cakes. If they took the food readily, it was a sign
of good luck for the year . 56 Here we have a pre-deistic snake
cult drawn into the orbit of the Delphic Apollo.
On the Hill of Kronos overlooking the sacred grove at
Olympia was a shrine of Eileithyia, the goddess of childbirth,
which housed a snake called Sosipolis, ‘saviour of the state’.
It was fed on honey cakes by a priestess, who alone was per-
mitted to enter, her head veiled . 67 The story was that, when the
men of Elis were about to join battle with the Arcadians, an
Elian woman set down between the opposing armies her new-
born infant, which at once turned into a snake and so terrified
the enemy that they took to their heels and fled. The snake
then vanished into the ground at the spot where the shrine
was afterwards erected . 58
E. N. Gardiner, in his monograph on Olympia, in which
he was careful to shield the classical Greek athlete from the
indignities of comparative anthropology, dismissed this cult
and the legend attached to it as ‘typical of the superstitious
credulity of the fourth century ’. 59 He forgot that an almost
identical cult flourished throughout the classical period in the
full blaze of the glory that was Greece on die Athenian
acropolis. In a famous passage of Herodotus we read that,
when the Persians were closing in on Athens, the sacred snake
6 ® All. NA. 11. 2. There was a similar cult at Lavinium: ib. 1 1. 16.
6 ? Paus. 6. 20. 2. This shrine has been identified as the Idaean Cave of
Pi. O. 5. 8 (C. Robert SO 41), suggesting a Minoan origin. The statue of
Sosipolis at Elis held the horn of Amaltheia (Paus. 6. 25. 4), which was a
Minoan symbol (p. 250 n. 10), and see further p. 292.
68 Paus. 6. 20. 4-5. For other snafce cults connected with childbirth see
Pi. O. 6. 45, Apld. 3. 6. 4. 59 Gardiner 125.
IV
Il6 STUDIES IN ANCIENT GREEK SOCIETY
kept in the shrine of Erechtheus mysteriously vanished, and so
reconciled the people to the evacuation of the city . 60 In this ___
case too the reptile was tended with monthly offerings of
honey cakes, and it was called the Housekeeper Snake,
because it was believed to have in its keeping the safety of the
state . 61 With it was associated the myth of Erichthonios, who
was a son of the earth goddess, or, in one version, of Athena
herself . 62 He is said to have been born as a snake or to have
been tended at birth by a pair of snakes . 63 His spirit was em-
bodied in the animal kept in the shrine, and from him was
descended the Athenian king Erechtheus, whose daughter
Kreousa, when she exposed her infant son, adorned it with a
snake necklace in his memory . 04
On this cult Nilsson remarks: 'It can be understood why
Athena is associated with the guardian snake
of the house if she originated in Mycenean
times. The goddess worshipped in the Minoan
domestic shrine was a snake-goddess.’ What
then was the origin of the snake-goddess? ‘It
has been thought’, he says, ‘that the snake
represents the soul of the deceased. . . . How-
ever, the snake is not always the representative
of the dead. Both ancient and modern folklore
know it as the protector of the house, and in
the Greece of our own day it is still called the
Lord of the House and receives offerings. There
is no need to look any farther for an explanation of
the Minoan domestic snake-goddess .’ 66 Athena,
hg. 2. Athena and Le argues, was associated with the shake in the
Sm Melos J r0m historical period because she had been associated
with the snake in the prehistoric period for the
same reason as the snake is called Lord of the House and
tended with offerings by the modern Greek peasantry. For what
reason? This explanation, beyond which we are not allowed to
look, explains nothing.
00 Hdt. 8. 41. 2-3. 01 Hsch. olKovpiv 691V. 02 Apld. 3. 14. 6; see p. 262,
63 Paus. x. 24. 7, Hyg. Ast. 2. 13, E. 21-3, 1427-g.
64 E. Io 18-26. * y
06 Nilsson HGR 26-7, 13, c f. MMR 283-4, Evans PM 4. 153-8.
IV . GREEK TRIBAL INSTITUTIONS H7
The value of modem Greek folklote for the study of ancient
Greece is undeniable, but surely, before skipping two mil-
lennia, we might have been permitted a glimpse at those
ancient sepulchral reliefs, discussed by Jane Harrison, in
which we see the deceased at a meal with a snake rearing
behind him or drinking from a cup in his hand. The snake is
fig. 3. Burial mound and snake: Attic vase
the dead man’s double. Then there is that black-figured vase,
also quoted by her, on which a snake rises from a tomb in
pursuit of a man retreating into the background . 00 Just as
Orestes was persecuted by his mother’s Furies in the shape of
snakes or snake-like women, so here the fugitive is evidently a
murderer pursued by the spirit of his victim. These serpentine
Erinyes were spirits of the dead. Again, thanks largely to
Nilsson’s own reasoning, it is agreed that the Greek hero cults
originated in the worship of the dead ; 07 and the heroes were
in the habit of appearing as snakes. After relating the death of
Kleomenes, who was saved from the vultures by a snake coiling
itself round the corpse, Plutarch adds that ‘the ancients
believed that the snake was associated with the heroes more
intimately than any other animal ’. 08
Another snake-hero, Kychreus, appeared in some of the
so Harrison PSGR 237, 325-31, T 267-71: figs. 3-4.
07 Nilsson HGR 103-4. 08 Plu. Cleom. 39.
Il8 STUDIES IN ANCIENT GREEK SOCIETY
IV
HG. 4. Feast of the dead: Laconian relief
Greek galleys at the Battle of Salamis. We are told that,
driven from Salamis, he had been received by Demeter at Eleusis,
where he remained in his animal form as her attendant . 69
Demeter too was a Minoan snake-goddess. This testimony
comes from the venerable Hesiod, who thus establishes the
very point which Nilsson dismisses with an appeal to modern
folklore.
The peasant customs of modern Europe are the mere
detritus of outmoded ritual, and so they usually need that
ritual to explain them. It would consequently count for very
little against the ancient evidence if in modern Greek folklore
the snake had completely lost its primitive significance. But
it has not. Unbaptised infants are popularly known as drdkoi,
00 Paus. 1. 36. 1, Apld. 3. 12. 7, D.S. 4. 72, Plu. Sol 9, Thes. 10, Hes.
fr. io7=Str. 393.
IV GREEK TRIBAL INSTITUTIONS II9
‘snakes*, because it is believed they are liable to turn into
snakes and vanish . 70 This is what happened to the Olympian
baby.
The direct evidence is decisive, but comparisons may still
be helpful, because it is always an advantage to view these
problems in the widest possible perspective. If, however,
we are going to venture beyond ancient Greece, why stop at
the Greece of our day? In regard to snake worship the Greeks
were at one with die ancient Egyptians and Semites and
primitive peoples in all ages and all parts of the globe. The
belief in snakes as incarnations of the dead belongs to the
common heritage of mankind . 71 The snake casts its slough,
thus renewing its vitality, and hence becomes a symbol of
immortality, of the power to be born again. That accounts
for its part in innumerable fables purporting to explain how
death came into the world and all our woe. In the Melanesian
languages the current phrase for ‘everlasting life* means liter-
ally ‘to cast the slough *. 72 In the Egyptian Book of the Dead the
deceased prays to become like the serpent: ‘I am the serpent
Sata. ... I die and am born again .* 78 The Phoenicians believed
that the serpent has the faculty not only of putting off old
age and renewing its youth but of increasing its strength and
stature . 74 And if by casting its slough it throws off old age, we
need look no farther for an explanation of the Greek word for
‘slough’, which was' gins, ‘old age’ (Latin senectus ).
The Zulus bury their dead in sacred woods, each of which
contains a number of cemeteries corresponding to the villages of
the district. The woods are taboo except to the priest, to
whom the dead frequently appear, sometimes as mammals but
usually as snakes . 78 Once, when the inhabitants of a kraal
were out celebrating a wedding feast, some old women,- who
had remained behind in one of the huts, were horrified to see
two snakes crawling along the wall. They sent for the village
70 Harrison PSGR 3 31, Polites P 2. 58, Demetrakes r.v.
vi Briffault 2. 641-51, 660-73.
72 lb. 2. 643.
73 Budge GE 2. 377. The same idea underlies the Greek legend of Glaukos
son of Minos (Apld. 3. 3. 1-2).
74 Eus. PE. 1. 10.
76 Junod LSAT 2. 376-7, 384-5.
120 STUDIES IN ANCIENT GREEK SOCIETY IV
headman, who was able to reassure them: ‘Don’t be afraid—
they are only our ancestor gods come to share the feast’. 76
Among the Masai, when a notable person dies, his soul goes
into a snake, which enters his kraal to look after his children. 77
A Greek would have said that he becomes a hero and guardian _
of the house. Every Masai family and clan has a particular
species of snake, which is believed to embody its ancestors.
Worsted in a fight, a man will summon his family snakes
with the cry: ‘Avengers of my mother’s house, come out!” 78
Such an appeal would have needed no interpreting for the
people brought up on stories of Clytemnestra and Sosipolis.
When these Bantus have won their liberty, they will make
good archeologists.
In snake-worship the clan totem has been replaced by a
generalised symbol of reincarnation. It is totemism in a
modified form.
6. Totemic Survivals: Clan Emblems
Returning to the Athenian acropolis, we have seen that in
the family of Erechtheus the snake appears both as ancestor
and as emblem. This comes near to saying that the Erech-
theidai were a snake clan. But they are not known to have
existed as a clan. The name was used as a poetical title for the
Athenians, and in historical times the cult itself was ad-
ministered by the Boutadai. 70 The evidence in this case is not
70 lb. 2. 384, cf. Hollis NLF 90. An old priest of the Bathonga des-
cribed to Junod how, on his entering a sacred wood to sacrifice, a snake,
the father of Makundju, came out, circled round him and his companions
and said, 'Thank you! so you are still there, my children! you come to load
me with presents'; and when Junod asked whether this was fact or fiction,
the old man replied, ‘Undoubted fact! These are great truths!' (LSAT 2.
384-5). The only important difference between the Greek and Bantu snake
cults is that the former were administered by women.
77 Hollis MLF 307-8, cf. Krige 53, 62, 65, 174, 285. The transition
from clan cult to state cult can be seen at the Baganda snake oracle on
Lake Victoria Nyanza: the oracle is public, but the office of interpreter
is hereditary in a particular clan: Roscoe B (191 1) 320-2.
78 Hollis MLF 308.
70 Harp. 'ETtopovriBcn, Paus. 1. 26. 5, Apld. 3. 15. x.
IV GREEK TRIBAL INSTITUTIONS 121
complete, and the most that can safely be said is that it pre-
supposes a totemic ideology.
The Spartoi of Thebes were so called because they traced
their lineage to the dragon's teeth sown by Kadmos at the
foundation of the city. 80 Epameinondas, the Theban leader who
fell at Leuktra in 362 B.c,, was buried in a tomb which bore
a shield emblazoned with a dragon in token of his membership
of this clan. 81 If the clan ancestor was a dragon and the clan
emblem was a dragon, the dragon must have been the totem
of the clan.
There was a Phrygian clan, called the Ophiogeneis, ‘snake-
bom’. It had a hereditary cure for snakebite and traced its
ancestry to a child begotten on a woman by a snake in a sacred
wood of Artemis. 82 Here a totemic myth of the normal type is
associated with a cult like the one we have noted in Epeiros.
All the great Attic clans had their ancestral emblems, which,
like the Spartoi, they displayed on their shields —
the trishlis of the Alkmeonidai, the horse of
the Peisistratidai, the horse’s hindquarters
of the Philaidai, the ox-head of the Boutadai,
and others not identified. 83 The trishle's is the
fylfoot or swastika, a symbol whose origin
is obscure. 84 The Peisistratidai traced their
pedigree through Peisistratos son of Nestor to
Poseidon, one of whose animals was the horse. 88
The horse’s hindquarters of the Philaidai were
evidently what is known as a ‘split totem’,
resulting from the division of a clan. 86 The
other half was probably the Eurysakidai.
Philaios and Eurysakes, the two sons of Ajax,
migrated from Salamis to Attica, where they
settled at Brauron and Athens respectively. 87 The ox-head has
been identified by its appearance on the coinage at the time
when the political influence of the Boutadai was at its height. '
80 Pi. P. 5. 101 sch., E. Ph. 94a sch., Paus. 8. n. 8. 81 Paus. 8. 11. 8.
82 Str. 588, JEl. NA. 12. 39. 88 Seltman 24, 30, 49, cf. Plu. Ale. 16.
84 Haddon EA 282. 86 Hdt. 5. 65. 4.
86 Frazer TE. 1. 10, 58, 77, 2. 397, 520, 536, 3. 100, 4. 175.
87 Plu. Sol. 10, St. B. <t>A«t8cn, Harp, EGpwdreiov, Paus. 1. 35. 3, Pher. 20.
122 STUDIES IN ANCIENT GREEK SOCIETY ' IV
It is probable, though not quite certain, that these ‘sons of
the oxherd 1 (boittes) had a hereditary part in the Athenian
Bouphonia, 88 a festival which shows clear signs of being
modelled on a communal clan feast. 89 It consisted of the
sacrifice of an ox, followed by a rite of expiation such as is
commonly employed for infringements of the totemic taboo. 80
Another Attic clan, the Euneidai, who held the priesthood
of Dionysos Melpomenos, were descended through Hypsipyle
of Lemnos from Dionysus, the god of wine. 91 Once, when
Hypsipyle was on the point of being put to death, she was
saved by the unexpected intervention of her sons, who proved
their identity by revealing the clan emblem, which was a
golden vine. 92 And lastly, the Ioxidai of Lycia, sprung from
Theseus, were forbidden to burn asparagus, which they
worshipped in memory of their ancestress Perigoune, who had
hidden herself in a bed of asparagus when pursued by Theseus. 93
Whether these traditions ‘necessarily demand a totemistic
explanation 1 is a question the reader must judge for himself,
remembering that those who reject this explanation have no
other. The last instance, in particular, in which the totemic
taboo has survived and the species is still worshipped in its
totemic form, might- seem to be incontrovertible. Frazer, who
had at least studied totemism, admitted that ‘this hereditary
88 The question turns on IG. z. 1656 UpJ»s BoOtou, which Toepffer
AG 159 takes to stand for lepcOs BoOrou, not kpEuj Bouttis, but the priest in
charge of the Bouphonia bore the title [BoOttis (Hsch. s.v.) and cf. IG. 3. 71,
294, where poujOyris is at once the priest’s title and the eponym of his clan
(Toepffer AG 136).
89 Robertson Smith R S 304-6.
00 FrazerTE 1. 18-20, 2. 156-8, 160, 3. 67, 81. The manner of selecting
the ox — it was induced to eat some com laid for it on the altar (Paus. I.
24. 4)— was designed to throw the responsibility on the animal: for
similar expedients see Paus. 2. 35. 6, Porph. Abs. 1. 25, Paton 83, cf. A. A.
1296-7. Its hide was afterwards stuffed with straw and yoked to a plough
(Porph. Abs. 2. 29— 3 °)» which suggests that the taboo that had been broken
may have been the ancient ban on the slaughter of plough-oxen: JEl VH. 5.
14, Arat. 132 sch, D.L. 8. 20, cf. Philost. lm. 2. 24.
01 3 * 2 74 > 2 7 &> Paus. 1. 2. 5, 1. 3 *» 6, Hsch, EOveTScn, cf. II. 7, 468—9.
They were professional lyre-players and dancers (Hsch., Harp., Phot, s.v.)
and had charge of the -nonuol or state processions (Poll. 8. 103): see p. 196.
02 A.P. 3. 10, E. fr. 765. 03 piu. xhes. 8.
IV GREEK TRIBAL INSTITUTIONS I 2 J
<►
respect shown by all the members of a family or clan for a
particular species of animal or plant is reminiscent of totemism’,
but added cautiously that ‘it is not necessarily a proof of it’. 94
So nice a distinction suggests that, where totemism is con-
cerned, the standard of proof is raised in proportion as the
evidence accumulates. In justice to Frazer it should be men-
tioned that, twenty-five years before making the comment just
quoted, he had expressed the opinion that ‘totemism may be
regarded as certain for the Egyptians and highly probable for
the Semites, Greeks, and Latins'. 98 In those days bourgeois
thinkers were less chary of general conclusions than they are
now.
7. Clan Cults and State Cults
Since there is no member of the Olympian pantheon who
is not associated in all sorts of ways with animals an4 plants,
it is a legitimate presumption that Greek religion in general
rests on a totemic foundation. A comprehensive study along
these lines would yield valuable results. Here I shall only
illustrate by a few concrete examples what I believe to have
been the fundamental process in the evolution of Greek re-
ligion — the transformation of clan cults into state cults, due
to the dissolution of tribalism and the rise of the city-state.
In the year 514 B.c. the Athenian tyrant Hipparchos was
assassinated. His assailants were two young noblemen, Har-
modios and Aristogeiton, who belonged to the Gephyraioi.
This clan was another offshoot from the stock of Kadmos. Its
first home on Greek soil had been Eretria (Euboia). From there
it migrated across f the straits to Tanagra (Bceotia). Expelled
from Tanagra after the Trojan War it settled in Athens, where
it maintained a secret hereditary cult of Demeter Achaia.
This we learn from Herodotus. 96 It is a clear case of a clan cult
surviving as such down to the fifth century.
Kadmos, whom the Greeks described as a Phoenician, reached
Thebes on his wanderings in search of his sister, Europa,
whom Zeus had ravished on the coast of Syria and carried off
to Crete. 97 Europa was the mother of Minos, the legendary
94 Frazer A 2. 125. 96 Frazer TE 1. 86.
96 Hdt. 5. 57, 61. 97 Apld. 3. 1. 1.
IV
124 STUDIES IN ANCIENT GREEK SOCIETY
king of Knossos. She is a parallel figure to Demeter, both
being emanations of the Minoan mother-goddess . 98 A cult of
Demeter Europa, founded presumably by the Kadmeioi, was
established near Thebes at Lebadeia," and at Thebes itself the
temple of Demeter Thesmophoros is said to have been the
palace of Kadmos . 100 It may be inferred that the Kadmeioi were
immigrants from Crete who brought with them a cult of the
Minoan mother-goddess. Their Phoenician ancestry will be
discussed in a later chapter.
At the end of his life Kadmos turned into a snake. The
Boeotian ships that sailed to Troy had figure-heads of Kadmos
carrying a snake . 101 It seems then that the Demeter" of the
Kadmeioi was a snake-goddess, like the Athena of the Erech-
theidai, and she survived among the Gephyraioi in their cult
of Demeter Achaia.
Demeter Achaia was also worshipped at Tanagra, Thespiai,
and Marathon . 102 Between Tanagra and Marathon lay Aphidna,
the birthplace of Harmodios and Aristogeiton . 103 Evidently
it was only one branch of the clan that settled in Athens.
At -Athens, according to Herodotus, the cult was strictly
confined to the clan. We know, however, from inscriptions
that in the fifth century a seat was reserved for the priestess of
Demeter Achaia along with other religious and civic officials
in the front row of the Theatre of Dionysus . 104 It may well be
that this was a privilege conferred on the Gephyraioi in
recognition of the part they had played in overthrowing the
tyranny. Their clan cult has thus been followed down to the
very point at which it is being taken over by the state.
Other instances reveal the actual transfer. When the Mes-
senians threw off the Spartan yoke in the fourth century, they
98 Nilsson MOGM 33, Famell CGS 2. 479, Roscher LGRM 1. 1417,
Persson 303-8, Picard PPD 336.
09 Paus 9. 39. 5.
100 Paus. 9 16. 5. The Theban Thesmophoria was celebrated in the
Kadmeia: X. Hell. 5. 2. 29.
191 Apld. 3. 5. 4, E. IA. 253-8.
10- Famell CGS 3. 323-4. Oropos, near Tanagra, was connected with
Eretria by a common cult of Eretreus (Str. 404) ^od a common dialect
(Buck GD 172).
103 Plu. M. 628d.
104 IG. 3. 373.
IV .GREEK TRIBAL INSTITUTIONS 12 5
reinaugurated the Mysteries of Demeter at Andania, their
ancient capital. The clan to which the cult had- belonged was
still in existence, and we possess the text of a decree in which
the clan chief, Mnasistratos, is appointed the first hierophant
under' the new regime and at the same time surrenders the
administration of the Mysteries to the state . 105
Another instance is furnished by an inscription from Chios,
where there was a phratry of six clans, the Klytidai, with a cult
of Zeus Patroios. At the time in question several citizens, who
did not belong to any of the constituent clans, had obtained
admission to the phratry. In Chios, as in Attica (p. To 7 ), the
phratry was ceasing to be exclusive. These citizens now claimed
the right to participate in the cult. It was decreed that a temple
should be erected to the god, and that the clansmen should
bring the sacra from their private houses to the temple on
certain feast days. This regulation was to come into force
immediately, and after an agreed term of years the sacra were
to be housed permanently in the temple . 1 08 The building of the
temple, which was doubtless paid for out of public funds,
marks the transfer from clan to state.
All over Greece we hear of priests holding office by right of
birth . 107 When the state took over, the clan usually retained
this ancient right. At Ithome and again at Aigion we hear of
a priest entitled to keep the god’s image in his private house
except when it was brought out for the annual festival . 108 We
infer that it had once belonged to his clan. At Halikarnassos
we hear of a cult comprising, in addition to the annual public
festival, a monthly service conducted privately at the new
moon . 100 This is a transitional stage in which clan cult and
state cult are combined. Other instances might be quoted of
the same kind. They show what Aristotle meant when he said
that it was characteristic of democracy to reduce the number
106 SIG. 736 n. 3, 9. Andania had been the seat of the Messenian kings
(Pans. 4. 3. 7); so this too may have begun as a palace cult, like the Theban,
and see p. 193.
io° SIG. 987.
107 JC. 12. 3. 514-9, 522, 865, 869, Stipp. Epig. Gr. 4. 282 etc., cf, PI.
Leg- 759 a -
xo8 Paus. 4. 33. 2, 7. 24. 4, cf. 9. 16. 5; Reinach TEG 141.
100 SIG. X015. 24, cf. Porph. Abs, 2. 16, Clem. Str. 3. 2.
IV
126 STUDIES IN AN CIENT GREEK SOCIETY
of cults and at the same time throw them open to the people . 1 10
The old famili es were not expropriated, but they were forced
to accept state control.
We must not of course suppose that these clan cults had all
maintained an unbroken history from tribal times. There were
constant struggles between rival clans for the political power
that went with religious administration. A single clan might
secure cults that did not belong to it or be forced to surrender
a share of its own. In these struggles our friends the Boutadai
had taken an active part. The oldest cults on the Athenian
acropolis were those of Athena Polias and Poseidon Erechtheus.
Both were administered by the Boutadai , 111 but they must
have been originally independent. Athena and Poseidon had
been rivals for the possession of the Acropolis . 1 12 The snake
in the shrine of Erechtheus belonged to Athena , 113 and, since
the hero embodied in it, Erichthonios, was Erechtheus’ grand-
father, the worship of both must have been included in the
palace cult of the Erechtheidai. The Boutadai, it is true,
claimed that their eponym, Boutes, was a brother of Erech-
theus and therefore also descended from Erichthonios ; 114 but
the Boutadai were interested parties, and in Hesiod Boutes is
a son of Poseidon, n 6 This gives the clue. After appropriating
the royal cult, which they combined with their own cult
of Poseidon, they confirmed themselves in possession by
affiliating their founder to the dynasty whose place they had
usurped.
The same sort of thing happened elsewhere. At Syracuse
the Mysteries of Demeter were hereditary in the clan of the
tyrant Hieron, who was an immigrant from Gela. They had
been brought to Gela by the clan ancestor, Telinos, from Telos,
an island off the promontory of Knidos, where there was
another cult of the goddess . 110 But the Syracusan Demeter was
110 Arist. Pol. 1319b. 19.
111 Apld. 3. 14. 8, PIu. M. 843b.
11_ ^ ee PP* 2.62-3. 113 Hsch. olKoupiv 6<piv. Apld. 3. 14. 8.
no Hes. fr. 101. There was yet another version making Boutes a grandson
or Ion (Apld. 1. 9. l6, Hyg. F. 14). This was probably the latest of the
three: see pp. 391-2.
110 7 * 1 5 3 " 4 > c f* Pi* O. 6. 92-5, 158 sch.; Famell CGS 3. 322.'
IV GREEK TRIBAL INSTITUTIONS 127
also known as Sit o, ‘Corn’, and as Simalis, which is not a
Greek word at a ill. 11 ’ This suggests that the pre-Greek popula-
tion of the city had worshipped a Sicilian corn-goddess, whom
Hieron annexed to his own Demeter when he made himself
master of the Syracusan state.
8 . The Clan Basis of the Eleusinian Mysteries
If we wish to uncover the tangled undergrowth of clan cults
in which the great panhellenic festivals had their roots, we
cannot do better than study the early history of the Eleusinian
Mysteries.
There were two clans in charge, the Eumolpidai and
Kerykes, with a subsidiary role assigned to the Krokonidai . 118
The founder, Eumolpos, was the common ancestor of the
Eumolpidai and Kerykes. It was natural that the clan most
prominent in the administration should have won pride of
place in the tradition, but in this tradition there are some
significant flaws.
The god of the Eumolpidai was not Demeter, to whom the
Mysteries were devoted, but Poseidon, and Eumolpos was a
stranger from Thrace . 119 From that direction he might easily
have brought a cult of Poseidon, who had many ties with the
north, but scarcely of Demeter, who cannot be traced in early
times further north than southern Thessaly . 120 The barbarous
ancestry of Eumolpos was evidently a source of embarrass-
ment to his descendants, because one authority assures us-
that the founder of the Mysteries was not the Thracian but
another man of the same name . 121 That this revised version
failed to establish itself is probably due to the presence in
the Mysteries of recognisable Thracian elements, notably the
117 Ath. 109a. 416b. The allusion is perhaps to the puXAot, cakes in
the shape of pudenda mulicbria, which the women baked for the Syracusan
Thesmophoria: Ath. 647a.
118 The Krokonidai (Toepffer AG 101-9, Deubner 75-7) were a crocus
'dan, associated with the threads of saffron ((96x05) which the mystics
wore on the right hand and foot (Phot, xpoxouv).
119 Apld. 3. 15. 4, Iso. iz. 193, Paus. 1. 38. z.
120 II. z. 695-6. 121 1st. 21*.
IV
128 STUDIES IN ANCIENT GREEK SOCIETY
name Brimo applied to Demeter or Persephone . 122 Ritual is less
plastic than myth.
The affiliation of the Kerykes to Eumolpos was not accepted
by the Kerykes themselves. They said that Keryx was a son of
Hermes by a daughter of Kekrops, the first king of Athens, 122
Kekrops takes us back four generations before die advent, as
traditionally dated, of either Eumolpos or Demeter. Hermes
too left his mark on the Mysteries. He was the consort of
Daeira, identified by Aeschylus with Persephone, in an ancient
form of the sacred marriage . 124 It would appear that the mystical
theogamy of Eleusis was older than the coming of Demeter.
Who brought her there?
At Athens she had been worshipped from early times as
Demeter Thesmophoros . 126 Herodotus says the ritual of the
Thesmophoria was brought from Egypt by the daughters of
Danaos, who settled in Argolis and there transmitted it to the
women of the indigenous Pelasgoi . 126 This is acceptable if we
may modify the historian’s well-known predilection for
Egyptian origins to the extent of interposing Crete between
Egypt and Greece.
There were traditions of her coming at several places in
Argolis — Argos itself, Hermione, Troizen. The Argives main-
tained that Triptolemos, the Eleusinian king who acquired
from her the art of tilling the soil and taught it to his people,
was really a son of one of her Argive priests who settled at
Eleusis , 127 Evidently they regarded the Eleusinian Demeter as
an offshoot of their own. The Athenians disagreed. They
acknowledged no debt to Argos. But the Arcadians did. Their
Demeter Mysia at Pellene was so named, they said, after one
122 Clem. Pr. z. 13, Tz. ad Hes. Op. 144, Ps. Ong. Philos. 170; Harrison
PSGR 551-3. Immarados, a son of Eumolpos (Paus. 1. 38. 3.), seems to stand
for *l<Ttiap< 5 toi 6 os, i.e. liiep&oiSos= 60 poXtros: it is also given as Ismaros (Apld. 3.
1 5 - 4 )« which was a mountain in Thrace (Od. 9. 198) near Maroneia (Str. 331.
44, cf. Od. 9. 196).
122 Paus. 1. 38. 3.
124 A. fir. 277, A.R. 3. 847 sch.; Lobeck 1212-5; see p. 173.
122 See pp. 220-2. iso Hdt. 2. 171.
127 Paus. 2, 18. 3, 2. 35. 4-8, 1. 14. 2. Her advent was also localised
at Pheneos (Paus. 8. 15. 3), Lakiadai (Paus. 1. 37. 2), Sikyon (Paus. 2. 5.
8), Kos (Theoc. 7. 5. sch.), and in Sicily (D.S. 5. 4.).
IV GREEK TRIBAL INSTITUTIONS I29
Mysios (the ‘mystic’?), who had welcomed her at Argos . 128
This is important, because the Eleusinian Demeter had ties
with Arcadia.
She_ was met at Eleusis by Metaneira and her daughters at
the Well of Flowers. 12 ® Metaneira was the queen, wife of
Keleos. One of her daughters married Krokon, eponym of the
Krokonidai, whose ruined palace was seen by Pausanias just
on the Eleusinian side of the old Attic frontier . 1 30 The Well of
Flowers lay on the other side of the town, on the road to
Megara . 131 This is the road the goddess would have come by if
she reached Eleusis from the Peloponnese.
Entering the Peloponnese from Megara, we come to
Phleious. Here, at a village called Keleai, were local mysteries
of the goddess, founded by one Dysaules, a brother of Keleos . 1 32
Keleos, Keleai — the connection is unmistakable. Keleos is
the male eponym of Keleai, just as Thespios is of Thespiai,
Alalkomeneus of Alalkomenai, Eleuther of Eleutherai . 133
Although king of Eleusis, he bears a Peloponnesian name. The
implication is that the cult he stands for was of Peloponnesian
origin. In the tradition as we have it, owing to the overriding,
prestige of Eleusis, the truth has been inverted. He is treated
as a native of Eleusis and his Peloponnesian connection is
explained by saying that the cult of Keleai was an offshoot of
the Eleusinian.
Keleai is a place-name of a common type. Like Thespiai,
Alalkomenai, Eleutherai, Potniai, Alesiai, it connotes a
women’s local cult, and it means literally the ‘crying women'
(kak'Oj kelomai ). 134 Once a month the village women go out
to the crossroads and cry to the moon. The custom is
128 Paus. 7. 27. 9, 2. 18. 3.
128 Horn. H. 2. 105-10, 161, • 184-7, 206-7, Paus. 1. 39. 1.
138 Paus. 1. 38. 1-3.
131 Paus. 1. 39. 1.
132 Paus. 2. 14. 1-4, Harp. AwbMtis. Keleai, which was the Homeric
Araithyrce (Paus. 2. 12. 5, II. 2. 571), cannot have been far from Pyraia,
where there was another cult of Demeter (Paus. 2. 11. 3). Dysaules was also
given as the father of Triptolemos: Paus. l. 14. 3.
133 D.S. 4. 29, Paus. 9. 33. 5, St. B. 'EteuBepcrf.
134 Cf. keXe 4 s 'woodpecker.' Kiazomenai, the 'screaming women,’ was
founded from Phleious: Paus. 7. 3. 9.
I
I30 STUDIES IN ANCIENT GREEK SOCIETY IV
CULTS OF DEMETER _ Map f I
IV GREEK TRIBAL INSTITUTIONS IJI
world-wide. 1 35 In Greece and Italy it was associated with
Artemis, and in later times with Isis, but above all with Demeter.
Servius describes the howling of Italian peasant women at the
■crossways in imitation of Demeter's search for her lost Per-
sephone . 136 There are other examples in this very dis trict.
At Megara there was a rock called Anaklethra, the rock of
‘invocation (anakaleo). Here Demeter had cried out for her
daughter, and the event was commemorated by the Megarian
women in a secret rite . 137 At Eleusis itself there was the
Laughterless Rock, where the goddess had sat down and
wept . 138 The ritual is not recorded, but it must have re-
sembled the Megarian , 139 and in it we have the clue to the
name Demeter Achaia, the Mourning Demeter (achos, ‘grief '). 140
In the hills west of Phleious lay the town of Pheneos. It had
two cults of Demeter. One was said to have been founded by a
descendant of Eumolpos. In the other, which is described as the
older, she was called Demeter Thermia , 141 which is a dialect
variant of Demeter Thesmophoros . 142 Its founder was Trisaules,
whose name recalls Dysaules, the brother of Keleos. Arcadia,
like Attica, was an ancient home of the Pelasgoi. Arkas, the
first king, is said to have introduced the art of agriculture,
which he had learnt from Demeter, exactly like Triptolemos . 143
One of his sons, Azan, took his name from the Azanes , 144
whose territory included Pheneos . 146 Another, Apheidas, was
136 Hastings s.v. Crossroads.
136 Apul. Met. xi. 2, Serv. ad Verg. A. 4. 609, E. 3. 26, cf. Horn. H. 2.
20-6.
I 3 ’ Paus. 1. 43. 2. At Athens (Eleusis?) the hierophant invoked Persephone
by beating an f|x rtov (Theoc. 2. 36 sch.); at Pheneos he smote the ground
with rods, wearing a mask representing Demeter (Paus. 8. 15. 2-3).
las Apld. x. 5. 1.
1 39 Cornford AEM 161.
140 PIu. M. 378c. A different explanation is given in EM. 'Ax<*tcc.
i« Paus. 8. 15. 1-4.
1 42 Cf. Kiptioi for K&iuoi (Buck GD 53) and Demeter Thermesia at Troizen
(Paus. 2. 34. 6).
143 Paus. 8. 4. 1.
144 p aus . 8. 4. 2 , Demeter was worshipped on Mount Azanion (Lact.
PI. ad Stat. Tb. 4. 292) and in Phrygia at a settlement founded by Azanes
from Arcadia (Paus. 10. 32. 3).
1 46 St. B. *A3avIo.
132 STUDIES IN ANCIENT GREEK SOCIETY IV
the eponym of Apheidantes, a village near Tegea. 140 Both re-
appear in Attica— Azan in the deme Azenia, 147 Apheidas in the
clan Apheidantidai. 148 What is more, their mother, the wife
of Arkas, is given as Metaneira, daughter of Krokon. 149
Herodotus was right. The cult of Demeter at Eleusis was in
origin a local form of the Thesmophoria, 180 introduced from
Arcadia by the Pelasgian Krokonidai, who had got it from the
Pelasgoi of Argolis. Of the three clans that worshipped her at
Eleusis the closest tie belonged to the one whose part in
historical times was the least conspicuous.
Our enquiry into the clan origin of the Eleusinian Mysteries
has thus enabled us to unravel the prehistory of one of the
most important cults in Greek religion. As a form of the
Minoan mother-goddess, Demeter may well -have come
ultimately from Egypt, with which Minoan culture had many
connections. 181 She reached Greece by two main routes, repre-
sented by Kadmos and Danaos. The first passed through
Euboia to Bceotia, the second through Argos to the Pelopon-
nese. In Attica they converged. Demeter Achaia came from
Boeotia, Demeter Eleusinia from the Peloponnese.
9. The Treatment of Homicide
Tribal society recognises two capital offences — incest and
witchcraft. Incest is violation of the rules of exogamy; witch-
craft is the misapplication by individuals of magic, which was
designed for the service of the community. 1 88 Both are punished
summarily by the community as a body. Other offences, in-
cluding homicide, are what we should call torts: redress lies
149 Paus. 8. 45. 1. 147 Polem. 65, Str. 398.
148 IG. 2. 785. Apld. 3. 9. x.
16° The Thesmophoria of Eleusis is said to have been founded by Tripto-
Icmos when he became king there (Hyg. F. 147). On the other hand,
the Athenian form, which closely resembled the Boeotian (cf. Paus. 9. 8. 1),
must have been influenced by the cult of Demeter Achaia, because the
loaves baked for the occasion (cf. p. 127 n. x 17) were called &xmfwn (Ath. 1 09c).
181 On the Egyptian connections of Danaos see pp. 379-80.
16 “ Diamond 280, Robertson Smith RS., 264, cf. Briffault 2. 568, Roscoe
BB (1923) 34, Gurdon 77.
IV
GREEK TRIBAL INSTITUTIONS
133
with the victim’s kindred. This is the procedure known as
primitive self-help. 1 * 8
In Attic law prosecutions for homicide were private suits
( dikai ), not public (graphal ) 164 The initiative rested with the
victim's household and phratry. 166 The terms for prosecution
and defence meant properly to 'pursue' (di 6 ko) and to ’flee'
(pheugo ) — the English ‘hue and cry’. 150
In early times, when manslaughter carried no moral stigma,
there was no discrimination between intentional and accidental
homicide. If the manslayer was unable or unwilling to offer
acceptable compensation, he was forced to flee the country.
This was no great hardship, because, wherever he went, he had
as a suppliant a compelling claim on the hospitality of any
stranger to whom he might appeal. In the Odyssey, just before
embarking for Ithaca, Telemachos is accosted by a fugitive,
who explains that he has committed murder and his victims’
kinsmen are on his heels. Telemachos takes him home without
a moment’s hesitation and entertains him there for as long as
he cares to stay. 167 In other cases the fugitive is not only enter-
tained; his host gives him a piece of land, and sometimes a
daughter into the bargain. 168 These customs imply a land sur-
plus. A chief who had more land than he could till was ready
to endow any stranger that came his way as a welcome addition
to his manpower.
In the choice between compensation and revenge and the
manner in which the latter was effected we recognise the
Iroquois practice. But in Homer one detail is missing. It is
clear that the victim’s kindred were under the obligation of
revenge, but there is no hint that the manslayer 's shared in his
liability. Why the Homeric poets were reticent on these
points need not be discussed now. The facts can be supplied
from other sources. There is post-Homeric evidence that the
clan was, or had been, responsible for the conduct of its
members. In 621 B.C., after failing in an attempted coup
d’/tat, Kylon took sanctuary with his followers at the altars.
They were seized and put to death by some men of the
168 Calhoun 62-7. 164 lb. 9. 166 D. 43. 57. 160 Calhoun 64.
167 Od. 15. 272-81, 508-46.
IV
134 STUDIES IN ANCIENT GREEK SOCIETY
Alkmeonidai, who thereby incurred a pollution so grave that
it was still being cast in their teeth two centuries later . 188 In an
inscription from Mantineia we read of some men fined for
murders committed in the sanctuary of Athena Alea, and it is
stipulated that, if the fines are not paid, the clans of the
guilty persons shall be excluded from the sanctuary for ever . 180
Even in later times, when the clan had fallen to pieces, the
principle of collective responsibility survived in the traditional
formula of public imprecation: 'If I break this oath, let me
perish, myself and my clan!’ 163 — or sometimes ‘myself, my
household, and my clan ’. 162 The same form of words is in use
among primitive peoples to-day . 163
These customs enable us to analyse one of the fundamental
elements in Greek thought.
The Greek for obtaining satisfaction for an injury, especially
homicide, was in Ionic tlsin latnbano , in Attic dlken diSko. The
Ionic form means to ‘take payment’, corresponding to the rule
of compensation described above. The Attic is based on the
same use of diSko as we have noticed in the term for prosecution.
Dike is used in Homer of a ‘way’ or ‘custom’, also a ‘judg-
ment’. Hesiod too applies it to a ‘judgment’, and to Justice
personified. In Attic it denoted primarily a private suit as
opposed to graphe , the abstract idea of justice being expressed
by dikaiosyne, formed from dlkaios, ‘just’. Other adjectival forms
are endikos, ‘just’, and ekdikos , ‘unjust’.
The root meaning of dike is ‘path’. It is cognate with delknymi
(Latin died) ‘point out’ or ‘show’ — to ‘show the way’. Dlken
diiko tind is therefore properly to ‘pursue a man along the
path’, to ‘chase him away’. Path-finding is an important thing
in the life of savages , 164 and in the Indo-European languages
the words for ‘path’ go a long way back . 166 To stray from the
beaten track was dangerous, and in early Attica a curse was
pronounced on those who refused to show strangers the way . 166
168 Th. i. 126, Hdt. 5. 71. 160 IG. 5. 2. 262, SIG. 9.
161 SIG. 37. x, 360. 50, 526. 40, Supp. Epig. Gr. 4. 58, Lycurg. Leo. 79,
cf. A.C. 1004, 1014. ’’ “
162 D. 23. 67. 163 Hutton 166.
164 Junod LSAT 2. 54-5. " 165 Moorhouse 123-8.
106 Diph. 62. Cf. Krige 214: 'It was the duty of all Zulus to show the
road when asked, and they could be fined if they refused to do so.'
IV GREEK TRIBAL INSTITUTIONS 135
Paths also form natural boundaries for holdings of land and
territorial divisions. Hence the widespread custom of deposit-
ing refuse on roads and crossways, and the apotropaic rites to
which it gives rise . 187
The transition from 'way' in the sense of 'path' to 'way' in
the sense of 'custom' is a straightforward advance from con-
crete to abstract. Similarly, with the law of homicide to guide
us, there is no difficulty in following the development from
'path' through 'vengeance' to the idea of punishment in
general. And it is an equally easy gradation from 'path'
through 'direction' to 'judgment'. This explains why judg-
ments are spoken of as 'straight' or 'crooked '. 188 A straight
judgment is etidikos, 'on the track'; a crooked one is ikdihs, ‘off
the track'. The metaphor enshrines the original meaning of the
word. And lastly the personification of Dike as goddess of
punishment or judgment leads to the formulation of the
abstract idea of justice.
So far we have been concerned with cases of homicide in
which the parties belong to different clans. What happened
when a man killed one of his own clan? The distinction
is vital. Clan solidarity was founded on collective production,
which meant that the individual was incapable of surviving
except as a member of a group; and even later, so long as
ownership was vested in the clan, clan kinship was of all ties
the most binding . 188 The penalty for killing a fellow clansman
was consequently as drastic as the crime was rare.
In early Germanic society similar customs prevailed. The
procedure in cases of manslaughter between clans is thus
described by Gronbech:
The kinsmen of the slain man appear in pleno as accusers. It is the clan of
the slayer that promises indemnity, the clan that pays it. It is the clan of the
slain man that receives the fine, and the sum is shared out so as to reach every
member of the group. 1 ™
Gronbech goes on to explain that homicide between clans ‘is
not a crime against life itself, nor even to be reckoned as
187 Hastings r.v. Crossroads.
188 II. 18. 508, 23. 579-80. Hcs. Op. 219-21, 250.
188 Cf. Smith and Dale 1. 296, Roscoe B (191 1) 12, B (1923) 5.
170 Gronbech 1. 55, cf. Tac. G. 21.
136 STUDIES IN ANCIENT GREEK SOCIETY IV
anything unnatural’. But homicide within the clan is a very
different matter.
From the moment we enter into the clan, the sacredness of life rises up in
absolute inviolability, with its judgment on bloodshed as sacrilege, blind-
ness, suicide. The reaction comes as suddenly and unmistakably as when a
nerve is touched by a needle. . . . When the curse has been uttered, when the
clan has renounced the condemned man by taking the oath whereby the law-
thing ‘swears him out' ...» the outlaw is dead. He is flung out from the'
life of men. 171
So in Greece. The man who shed a kinsman’s blood was
hounded out of the community, pursued by the curses of his
kindred, or, as they expressed it, by his victim’s avenging
spirit, which ran him down and devoured him till he was
nothing but a heap of bones . 172 The curse and the avenging
spirit are the same thing. The Arai or Erinyes symbolise the
collective imprecation of the clan calling on the souls of its
ancestors to rise and destroy the outcast . 173 Accordingly he
went mad; or rather he was already mad when he did the deed.
What he has done is so fearful, so unheard-of, that it is proof
in itself of his incapacity to behave as a normal member of
society. Many instances are on record of savages actually dying
of horror at the discovery that they have violated uninten-
tionally some peremptory taboo . 174 The primitive conscious-
ness, being less complex than that of civilised man, is more
171 Gronbech 1. 343. So in Wales, for the murder of a kinsman ‘there
is no slaying of the murderer, ... no galanas [fine], nothing but execration
and ignominious exile’ (F. Seebohm TCAL 42). This distinction, which
was general in Celtic and Germanic law ( ib . 63, 66, 164, 166), explains
why in early medieval England ‘the homicide of a kinsman is still generally
free from judicial interference or criminal law: he is handed over to the
Church and his punishment is spiritual penance’ (ib. 336, cf. Gronbech
»• 35 )*
172 A. E. 244-66.
173 A. E. 420.
174 A native of Melanesia, asked how he would feel if he had committed
incest, replied, We don t do it; if a man did it, his mind having turned
wrong and silly, he would wake up and kill himself' (Wertham 179,
cf. Malinowski SRSS 95). It is widely believed that violation of the totemic
taboo results in insanity or leprosy or both (Frazer TE 1. 16-7) and instances
arc on record of natives refusing food and dying after receiving superficial
spear wounds, simply because they believed the spear to have been bewitched:
Spencer A 403-4.
IV GREEK TRIBAL INSTITUTIONS I37
easily deranged. And so the crime is its own penalty. He was
mad to do it; and if he regains his sanity, the shock of realisa-
tion drives him mad again.
This is the psychology that inspired the Greek concept of
ate — the fatal delusions inflicted by the Erinyes. In the cul-
tured milieu of the Homeric poems the word was largely
purged of its savage content, denoting in general little more
than a state of mental aberration leading to a disastrous
blunder. In Cretan Doric, on the other hand, it survived
simply as a legal term for penalty or damages . 178 In the one case
it is the subjective aspect, in tne other the objective, that has
become dominant to the exclusion of its opposite. But the
primitive unity is preserved by Aischylus, who applies the word
both to the sudden brainstorm that causes the crime and to
the self-destruction that is its consequence . 176
10. The Law of the Heiress
Some remarks by Dikaiarchos, a pupil of Aristotle, on the
evolution of the Greek tribal system have been preserved in a
Byzantine paraphrase. They show how the process was in-
evitably misinterpreted when viewed in the light of precon-
ceptions drawn from class society.
The clan (patra ) is one of the three Greek social units known as clan,
phratry, and tribe. When the group of kin, confined originally to the
married couple, was extended to the second degree, there arose the unit
called the clan, which was named after its oldest or most influential member,
c.g. Aiakidai, Pclopidai. The phratry came into existence because daughters
were given in marriage to another clan. The bride ceased to take part in the
religious life of her father’s clan, because she was included in her husband's,
and therefore, to replace the severed union between brother and sister,
another religious union was instituted, the phratry. And so the phratry arose
from the rclauonship between brothers just as the clan had arisen from the
relationship between parents and children. The tribe evolved from the
process of fusion into cities and nations, the components of which were
termed tribes. 177
178 Lex Cart. 11. 34.
176 A. A. 396-7, C. 270-1, 381, 595-6, E. 379-81. In my edition C.
383-4 has been mistranslated: it should be 'by mans of a reckless, criminal
hand.'
177 Die. 9.
IV
138. STUDIES IN ANCIENT GREEK SOCIETY
Dikaiarchos starts, like his master, from the premiss that the
primeval unit of society was the married couple. Aristotle had'
explained how in his view this unit had expanded into the
family, village, and city. 178 Here Dikaiarchos is applying the
same reasoning to the tribe. The premiss is of course a false one,
and so he has considerable difficulty in squaring it with the
facts. The facts themselves, however, are stated correctly.
The dan is an organic unit within the phratry, which is a
group of intermarrying clans. It has been left to modern
historians to falsify the facts in conformity with the premiss.
He is not thinlung primarily of the Attic system. That is
shown by his use of the term patra in place of the Attic ginos and
also by his statement that the bride became a member of her
husband's clan. And he says that ‘daughters were given in
marriage to other clans'. This testimony is specially valuable,
because in historical Athens the rule of exogamy had entirely
disappeared.' Its former existence can, however, be inferred
from the laws of inheritance, which define the circumstances in
which it might be infringed. But first let us consider an analogy.
The Semitic peoples were originally matrilineal, 179 but the
Jews, when they took to agriculture after settling in the land of
Canaan, were already patriarchal. All property, real and personal,
was transmitted in the male line. Land was inalienable; acquired
goods were distributed among the sons. But what if there were'
no sons? In the Book of Numbers we read (xxvii. 8):
If a man die, and have no son, then ye shall cause his inheritance to pass
unto his daughter.
This meant that the usufruct passed to the man she married,
who under the rule of exogamy would belong to another clan. *
Accordingly it was enacted (xxxvi. 8):
And every daughter that possesseth an inheritance in any tribe of the
children of Israel, shall be wife unto one of the family of the tribe of her
father, that the children of Israel may enjoy every man the inheritance of his
fathers.
The family of the tribe is the clan. 180 The heiress was compelled
to marry into her own clan in order to keep the property in the
male line.
179 Arist. Pol 1252. 179 Robertson Smith KMEA 27-34. 180 U. RS 276.
IV GREEK TRIBAL INSTITUTIONS 139
In Attic law, as codified in the sixth century, the sons in-
herited the paternal estate on condition that they dowered
their sisters at marriage. The dowry was the daughter’s share
of the inheritance. If there were no sons, the daughters inherited
the whole, but they could then be claimed in marriage by
their father’s next-of-kin . 181 At Gortyna the rule was the same,
except that the daughter inherited a share in her own right,
though smaller than the son's, and the heiress could refuse to
marry the next-of-kin by surrendering to him part of the inheri-
tance . 182 Though later in date, the Gortynian procedure is
more archaic than the Attic, and both rest on the same
principle as the Jewish. The rule of exogamy, and with it the
woman’s liberty, had been sacrificed to the male interest in
private property . 183
To recapitulate: the Greek tribal system resembles that of
the Romans and the Iroquois, similar in structure, origin, and
development. The otkos was evolved within the genos, as the
familia within the gens, by the growth within die clan of
smaller units, which eventually became independent. The
subsoil of Greek religion consists of totemic clan cults. The
principle of clan solidarity, which we have traced in parallel
customs among Greeks, Germans, and Amerindians, underlies
the terminology and procedure of Greek criminal law. The
evidence for exogamy is indirect, but none the less certain,
because exogamy is inherent in the structure of the clan.
There remains the mode of succession and descent. This will
occupy us continuously in the ensuing chapters. With it is
bound up not only the problem of the pre-Hellenic cultures of
the AEgean but also the distinctive character of Hellenism.
The present chapter may be concluded with some general
considerations which will help us to approach these larger
questions from the correct point of view.
181 The heiress was obliged to marry the next-of-kin as soon as she came
of age (Is. 6. 14), and die next-of-kin, if already married, divorced his
wife in order to marry the heiress (Is. 3 . 64). Strictly she was not an heiress at
all, but merely an appendage to the estate (frrtoaipos).
182 Lex. Cert. 4. 31-44, 7. 54-8. 6.
183 Plu. Sol. 21: ‘The property must remain in the glnes of the deceased.’
IV
140 STUDIES IN ANCIENT GREEK SOCIETY
11. Ancient Creek Ethnology
Surrounded as they were by more backward peoples at
various stages of savagery or barbarism and by the advanced but
archaic empires of the Near East, the civilised Greeks did not
fail to observe that the status of women in these surrounding
countries was very different from what it was in their own.
Their reports and comments on this subject are of great in-
terest. In point of accuracy they are of course open to question
except where they can be confirmed from other sources, and
most classical scholars have tended to discount them as
credulous travellers* tales. Modern ethnologists have been
more respectful. But, apart from their accuracy, they are im-
portant, because they reveal the form of words traditionally'
employed to describe primitive institutions at a time when
there was no science of ethnology.
One of our earliest informants is Herodotus. A native of
Asia Minor, he had travelled widely, not only in Greek lands,
but in Africa, Egypt, Syria, Mesopotamia, and the countries
bordering on the Black Sea. Admittedly he is not always
reliable, but some of the things he tells us are in close accord
with what we learn from other authorities, ancient and
modern.
The Agathyrsoi of Scythia, so he says, ‘have wives in com-
mon, so that they may all be as brothers to one another
without hatred or jealousy*. 184 The idea that community, of
wives goes with community of property was a familiar one.
We meet it again in Aristophanes and Plato. 186
Another Scythian tribe, the Galaktophagoi, are described
by Nicolaus .of Damascus as ‘having both property and wives
in common, and they call their seniors fathers, their juniors
sons, and their coevals brothers’. 188 This sounds like Type I of
the classificatory system (p. 60). Among the Geloi, also
Scythians, the women, according to Eusebius, ‘till the soil,
build the houses, do all the work, and lie with any man they
like without the reproach of adultery*. 18 ’ The sexual liberty of
the women is complementary to their activity in the labour
184 Hdt.4. 104. 188 Ar. PL 5 1 0-626, PI. 2 t. 41 6d. '
188 Nic. Dam. 123. i 87 Bus. PE. 9.
IV GREEK TRIBAL INSTITUTIONS I41
of production. Some of the Upper Libyans, says Aristotle,
'have wives in common'. 188 Details are furnished by Herodotus.
Among the Machlyes of Libya, he says, 'sexual intercourse is
promiscuous: they do not live together but copulate like
cattle'. How many missionaries have exclaimed in horror at
this abomination! When the children reach a certain age, 'they
are assigned to the men at an assembly according to their like-
nesses'. 188 This point may be fanciful, but it is obvious that in
such conditions there could be no concept of individual
paternity. Among the Nasamones of Libya 'sexual intercourse
is promiscuous, as among the Massagetai: they just set up a
staff in front of the hut and then they copulate’. 180 Here again
modern analogies are only too plentiful. 'When a man marries,
the bride is required on the first night to lie with each of the
company in turn’. 181 This practice, which we have met already in
contemporary Australia (p. 66), is so widespread that it has
been given die name of nasamonism in allusion to Herodotus.
It is only another form of the pre-nuptial promiscuity which he
records of the Lydians, Cypriotes and Babylonians and Plautus
of the Etruscans. 102 Again, the Massagetai of Central Asia, he
says, 'have wives in common. . . . When a man desires a woman,
he hangs his quiver in front of her waggon and enjoys her
there and then.’ 183 The Massagetai, or Great Getai, are known
from Chinese sources, which show that they were matriarchal,
and some authorities would identify them with the Jats of
Hindustan. 104 If this is correct, their customs must be assigned
to the prehistory of our own culture.
Strabo, who was well acquainted with' Asia Minor, Egypt,
Italy, and the western Mediterranean, says that the Cantabri
of Spain 'have a form of matriarchy ( gyttaikokratfa ); the
daughters inherit and give their brothers in marriage’. 106 The
same rule of inheritance obtained in ancient Egypt. 186 The
Ethiopians, according to Nicolaus, ‘hold their sisters in great
honour, and their kings are succeeded by their sisters’ sons,
188 Arise. Pol. 1262a. 9. 180 Hdt. 4. 180. 5-6, cf. Nic. Dam. ill.
100 Hdt. 4. 172. 2. 181 Hdt. l.t.
108 Hdt. 1. 93. 3, 1. 199, Plaut. Cist. 2. 3. 20. 183 Hdt. 1. 216.
181 Briffault 1. 334-9, Russell 3. 223-6. For other views scc'Tarn Si.
186 Str. 165. 188 See pp. 139-60.
IV
142 STUDIES IN ANCIENT GREEK SOCIETY
not their own '. 107 This is confirmed by Egyptianjmnals of the
Ammonian Dynasty . 108 In Lycia, says Herodotus, ‘if you ask
a man who he is, he replies by naming his mother and his
mother’s mother ’. 100 Among the Etruscans, according to .
Theopompos, the men ‘had wives in common and die
children did not know their own fathers ’. 800 That the Lycians
and Etruscans were matriarchal has been proved by archaeology.
These quotations show that what was meant by ‘having
wives in common’ was some form of group-marriage combined
- with common ownership, and that, when children are de-
scribed as ‘not knowing their own fathers’, the reference is to
descent in the female line. The Greeks were well acquainted
with the realities of primitive society.
It is in this light that we must interpret a tradition con-
cerning Kekrops, the first king of Athens, who was credited
with the invention of matrimony. Before his time ‘there had
been no marriage; intercourse was promiscuous, with the
result that sons did not know their fathers nor fathers their
sons. The children were named after their mothers ’. 201 So
matrilineal group-marriage had once prevailed at Athens.
There is no reason to discredit this tradition. Athenians
would not have fabricated a story which represented their
ancestors as savages. ‘The Greeks lived once as the barbarians,
live now.’ In these memorable words Thucydides enunciated
with characteristic insight the principle of the comparative
method in social anthropology . 208 The same truth is implicit in
the writings of ASschylus and Hippokrates . 203 That was the
materialist tradition. But already, in the time of Thucydides,
the reaction had set in. The materialist view of social evolution
was irreconcilable with the doctrine, fostered by the growth of
slavery, that Greek and barbarian were different by nature.
If such things as' primitive communism, group-marriage, and
matriarchy were admitted into the beginnings of Greek
civilisation, what would become of the dogma, on which the
ruling class leant more and more heavily as the city-state
107 Nic. Dam. 142. ids Revillout 2. 147.
100 Hdt. 1. 173. 5. soo Theop. 222.
201 Clearch. 49, Charax 10, Io. Ant. 13, cf. Vatr. ap. Aug. CD. 18. 9.
“° 2 Th. 1. 6 . 6 . S03 Q. Thomson AA 218-9.
IV
GREEK TRIBAL INSTITUTIONS
143
declined, that its economic basis in private property, slave
labour, and the subjection of women rested on natural justice?
If the writings of the later materialists, Demokritos and
Epicurus, had not perished, we might well have possessed
a more] penetrating analysis of early Greek society than
Aristotle’s. But they perished partly for that reason. Plato
wanted the works of Demokritos to be burnt , 204 and his wish
has been fulfilled.
No serious student can read Aristotle’s Politics without ad-
miration for the author's erudition and insight. If that book
had perished, the world would be the poorer. But this must
not prevent us from recognising its limitations. He knew that
the Greeks had once lived in tribes, and he must have been
famili ar with the tradition that they had once been without
slaves . 206 He was presumably aware of the part assigned to
Kekrops in the history of matrimony, and in any case he had
before him the example of contemporary Sparta, where the
rule of monogamy was so little binding that half a dozen brothers
might share a wife between them and adultery was not punish-
able or even discreditable . 2 06 Yet, accepting the city-state as
the only possible foundation for civilised life, he constructs a
theory in which the original nucleus of society is identified as
the married couple dominated by the male and supported by
slave labour. 2 07 The principle laid down by Thucydides was
precluded from the start.
Where Aristotle failed, we cannot expect much of Herodotus.
During all his travels the truth stated so lucidly by Thucydides
never dawned on him. All he has to say of the Egyptian
matriarchate is that 'sons were not obliged to support their
parents, but daughters were' 208 — alluding to the rule of in-
heritance; and the remark occurs in a passage where he is more
concerned to divert his readers than to interpret the facts.
Hence it is not surprising that he introduced his account of
204 Aristox. 83. 206 Hdt. 6. 137. 3.
206 The marriage custom is an instance of 'fraternal polyandry' (see p.
71). The children were treated as common: PIb. 12. 6. 8. Conversely, ,
the wives of different husbands might be temporarily exchanged: Plu. Lyc.
1 5, X. KL. 1. 9. Wife-lending was also an ancient custom at Rome: Str. 515,
Plu. Cat. 25, cf. App. BC. 2. 99, Quint. Inst. Or. 3. 5. 11, 10. 5. 13.
207 Arist. Pol. 1252b. 20s Hdt. 2. 35. 4.
IV
144 STUDIES IN ANCIENT GREEK SOCIETY
the Lydan matriarchate with the observation that ‘it is un-
paralleled among the peoples of mankind '. 209 The wish was
father to the thought. The significance of this misstatement is
that it represents what, for reasons that will appear in due
course, the Greeks of his day were predisposed to believe.
12. Linguistic Evidence of Matrilineal Descent -
In the late eighteenth and nineteenth centuries, when the
capitalist class viewed past and future alike with complete
self-confidence, a long succession of progressive thinkers —
Adam Smith, Ferguson, Millar, Bachofen, Morgan, McLennan,
Tylor — did much to correct the traditional self-portrait of
ancient society to which Plato and Aristotle gave the finishing
touches,* but, as capitalism moved into its decline, it became
apparent that such matters as private property and the status of
women touched the same prejudices in the modern bourgeoisie
as we have noticed in Aristotle; and in the present century
these prejudices have become even more sensitive, because
more irrational, through the abolition of private property and
of social inequalities between the sexes in the Soviet Union.
And so for the second time in history these aspects of ancient
civilisation have been expunged. There are of course exceptions.
Evans, Ridgeway, Harrison, Glotz, Briffault, and others have
insisted on the matriarchal character of prehistoric Greece.
But, apart from Briffault, who turned eventually to Marxism,
these were all primarily archaeologists, on whom a materialist
attitude was forced by the nature of their subject. And even
so, while recognising the truth, they cannot be said to have
appreciated its significance. Among the general run of historians,
and of course in the charmed circle of ‘pure scholarship’, the
matter is not discussed. ‘Democracy’, says Rostovtzeff in
his History of the Ancient World , ‘banished woman from the
street to the house ’. 210 The fact is noted, but with no attempt
to explain why democracy was so unchivalrous; rather, it is
taken for granted that democracy put her where she ought to
have been all along. And the Cambridge Ancient History is silent
even about the fact.
209 Hdt. i. 173.^4.
210 Rostovtzeff 1. 287.
IV GREEK TRIBAL INSTITUTIONS 145
Words are great telltales. They are speaking witnesses to the
vanished past.
The typical Greek clan name has the patronymic termination
- (das (-{des), based on the element -id-, which in Greek is
feminine . 211 It follows that in early times the women, and
not the men, had been regarded as representatives of the
clan.
The Greek adelphis and adelphi, 'brother* and 'sister*, are
without parallel in the other Indo-European languages . 212 IE
*bhrdter and *su/s5r survived in Greek as phrdter and /or, but
not as terms of relationship. The displacement of these terms
is the most distinctive feature of the Greek terminology, and
demands an explanation.
Phrdter denoted a fellow member of the phratry. At Athens,
when a boy came of age, he was admitted to his father’s phratry
at the Apatouria, the feast 'of the sons of the same fathers *. 213
In what sense were the phrdteres ‘brothers' and 'sons of the
same fathers'?
At Sparta, where the boys were enrolled in sodalities called
ag/lai, 2 ^ the term kdsios, ‘brother*, was applied to all brothers
and cousins in the same ag/la, 218 and in the form basis or bases it was
used to denote men belonging to the same generation as the
speaker . 213 The conclusion is clear. The Attic-Ionian phrdteres and
the Dorian kdsioi were originally, in each generation, the sons
of the same father, the sons of the father’s brothers, the sons
of the father’s father’s brother's sons, and so on. They were
‘brothers’ in the classificatory sense.
tor survives only in a late Greek lexicon, where it is ex-
plained in one entry as ‘daughter or cousin’, in another as
‘relative ’. 217 The explanations are obviously inaccurate and
211 Meillet GCLC 390-1, Buck CGGL 341, Chadwick, HA 359. The
Sema Naga patronymic has been analysed by Hutton 13 1-2 with a similar
result.
212 This account of dS&<p6s is based on Kretschmer GBB. '
213 A. Mommsen 323-49. The third vowel presents a difficulty: Deubner
232. Could 'Awrrotipiec have been an epic form (determined by metre, like
Trov/XuB6T£ipcc) on the analogy of po0voj=p6vos (*u6vfos)?
214 Plu. Lyc. 16-21. 218 Hsch. K¬.
210 Hsch. K&ffus, for which k&ctis should perhaps be read.
217 Hsch. top, tope;.
K
146 STUDIES IN ANCIENT GREEK SOCIETY IV
confused, but the original meaning of the word' is not in
doubt, because it is the Greek equivalent of IE *suesor.
The terms adelphSs and adelphd are adjectival; that is, they are
properly descriptive epithets, standing for phrdter ddelphos and
dor adilphe. The meaning of adelphSs is horn from the same
womb *. 218 Thus, phrdter ddelphos is a uterine brother as opposed
to phrdter Spatros, a ‘brother by the same father.
These words tell their own story. Drawn into the orbit
of the older Cretan and Anatolian cultures, the Greek-
speaking invaders of the Aegean adopted matrilineal succes-
sion, and the new application of the terms for brother and
sister was marked by descriptive epithets, which eventually
supplanted them. The men, however, retained the patrilineal
phratry, and in this connection phrdter survived. The women
had no corresponding organisation, and so /or disappeared.
The linguistic data are completely explained on this hypothesis,
and on any other they are unintelligible.
The meaning of the Greek adelphSs belongs to the domain of
what ‘every schoolboy knows’; yet how many of them have
been encouraged to enquire why the brother should have been
described as one ‘of the same womb’? There are even profes-
sional scholars who have never given it a thought. The begin-
ning of wisdom is an enquiring spirit. As they are continually
reminding us, intellectual curiosity was one of the virtxies of
the Greeks. It cannot be said to flourish under the meta-
physical ‘discipline’ of a classical education.
218 Kuiper 287 treats Kaalyvrvros as analogous in form to Lat. cognatus
and in meaning to &SeX<p6s, i.e. 'der mit einem zusammen geboren ist’: k&ctis
would then be hypocoristic. But it seems more probable that Kaalyvirro;
meant originally consobrinus, as opposed to dS&qxSs, i.e. ‘brother’ in the
classificatory sense: Suid s.v., II. 9. 464, 16. 456.
Part Two
MATRIARCHY
In eurem Namen, Mutter, die ihr thront
Im Grenzenlosen, ewig. einsam wohnt,
Und doch gesellig. Euer Haupt umschweben
Des Lebens Bilder, regsam, ohne Leben.
Was einmal war, in allem Glanz und Schein,
Es regt sich dort; denn es will ewig sein.
GOETHE
V
THE MATRIARCHAL PEOPLES OF THE AEGEAN
l. What is Matriarchy?
A MAJOR factor in man’s differentiation from the animals was
protraction of his period of growth, during which he was sus-
ceptible to formative instruction. This was imparted by the
females, whose maternal functions, in the absence of economic
production, necessarily placed them in control of the group.
The only distinctive function exercised at this stage by the
males was procreation. The group's habits, norms of behaviour,
inherited traditions, which constituted in their totality the
nucleus of human culture, were formed and transmitted
by the women . 1
The subsequent conflict between the sexes resulted, as we
saw in Chapter I, from the development of production. Under
a hunting economy there arose a contradiction between the
economic role of the males and their social status; and this
tendency, reinforced by stock-breeding and warfare — both
offshoots of hunting — effected eventually, where it operated
freely, a reversal in die position of the sexes. That is why, in
modern non-agricultural tribes, the matrilineal rule has been
overthrown by nearly fifty per cent of the hunters and all the
pastoralists . 2
It has been urged against Morgan that matrilineal descent
does not necessarily mean that society is controlled by the
women. This is quite true. We must distinguish between
matrilineal and matriarchal. In many, perhaps most, of the
matrilineal tribes known to us the actual control is in male
hands. The rule of succession itself is often circumvented by
transparent expedients, as when a man names his sons into his
1 Briffault i. 96-110, 195-267.
2 Hobhouse etc. 152 cite the Navahos of Arizona as pastoral and
matrilineal, but from Frazer TE 3. 242 it seems doubtful whether they
should be classed as pastoral.
V
150 STUDIES IN ANCIENT GREEK SOCIETY
own clan, or makes over his acquired wealth to them as a gift
before he dies . 3 In the spirit of subterfuge primitive man
shows all the ingenuity of a modern jurist. Change is justified
by pretending that things are as they were before.
The factors making for the supremacy of the male were off-
set by the discovery of agriculture. In contrast 'to hunting and
stock-breeding, which are both nomadic occupations, agricul-
ture prepares the way for one of the most momentous steps
in the whole record of human progress — the adoption of a
sedentary life. It was only after he had learnt to till the soil
that man could become in the full sense of the word a ‘political
animal’ — an animal that lives in towns . 4 This was the step
the Iroquois were about to take when they were interrupted by
European conquest, thus falling short of the Aztecs, who owed
their pottery, metallurgy, and architecture, their pictographic
script and lunisolar calendar, to the advance from nomadic
to sedentary agriculture. In the Old World the contrast is even
more striking. Some parts of the Eurasian steppe-land have
only become civilised in our own generation, while the rich
alluvial valleys of southern Asia have witnessed from time
immemorial the rise and fall of empires. The urban civilisa-
tions of the Nile, Euphrates, and Indus, which drew their
wealth from the soil, had their beginnings in the fourth mil-
lennium B.c., whereas the intervening deserts have remained
down to our own day the home of ‘such as dwell in tents and
have cattle ’. 6 There is no need to insist on the supreme im-
portance of agriculture. The point is that this mode of pro-
duction was initiated by women, who thus played the decisive
part in the origin of civilisation.
What then is matriarchy? In answering this question we
begin, in accordance with our method, by seeking in the
ethnological domain a living example of a matriarchal com-
munity. But here we are confronted with a difficulty inherent
in the nature of our search. The optimum conditions for the
survival of matriarchy would be a rapid advance from food-
gathering to agriculture. But these are the optimum conditions
for the development of civilisation. The object of our search is
3 Frazer TE I. 71, 3. 42, 72, 308; 2. 195, 3. 245, 4. 290.'
4 Arise. Pol. 1253 a. 9. 6 G «j. 4. 20.
V MATRIARCHAL PEOPLES OF THE iEGEAN 151
frustrated by the conditions necessary for its attainment. This
explains why so few examples of the matriarchate survive to-
day. It lies buried beneath the civilisations erected on it.
What we are looking for is most likely to be found in
regions where a rapid advance to the upper stages of barbarism
has been followed by stabilisation. Such regions exist in the
south and south-east of Asia. I quote from Marx:
These small and extremely ancient Indian communities, some of which
have continued down to our own day, are based on possession in common of
the land, on the blending of agriculture and handicrafts, and on an unalterable
division of labour, which serves, whenever a new community is started, as a
plan and scheme ready cut and dried. . . . The simplicity of the organisation
for production in these self-sufficing communities that constantly reproduce
themselves in the same form, and when accidentally destroyed spring up on
the same spot with the same name — this simplicity supplies the key to the
secret of the unchangeableness of Asiatic societies, which is in such striking
contrast to the constant dissolution and refounding of Asiatic states and the
never-ceasing changes of dynasty. The structure of the economic elements of
society remains untouched by die storm-clouds of the political sky.e
The Khasis are a people of some 200,000 souls inhabiting
the hills to the north-east of Dacca on the borders of Bengal
and Assam. Culturally they are isolated. Their language
belongs to the Kolarian family, represented by the Santhals and
Mundas of Chutia Nagpur and the Satpura Hills in the
Central Provinces. Their staple industry is agriculture, sup-
plemented by hunting, fishing, and stock-raising. The prin-
cipal crop is rice. Manuring is well understood, but in most
parts of the country the plough is unknown.
About half of the Khasi country is divided into minute
native states; the remainder belongs to British India. British
rule, direct or indirect, dates from 1835, and since 1842 the
population has been served by a Welsh missionary college,
with the happy results noted by Lieut.-Col. Gurdon, to whom
we are indebted for a valuable monograph. ‘Khasis who have
become Christians', he tells us — their number is over 20,000 —
‘often take to religion with much earnestness . . . and are
model Sabbatarians, it being a pleasing sight to see men,
s Marx K. 3 50-2. The stability of Indian society is attested by ancient
Greek accounts: E. R. Bevan in CHI 1. 391.
V
152 STUDIES IN ANCIENT GREEK SOCIETY
women, and children trooping to Church on a Sunday morning
dressed in their best with quite the. Sunday expression on their
faces that one sees in England / 7 Along with these benefits the
Khasis have managed to preserve their native customs, whose
importance may be judged from the following remarks by
Lyall:
Their social organisation presents one of the most perfect examples still
surviving of matriarchal institutions, carried out with a logic and thorough-
ness which, to those accustomed to regard the status and authority of the
father as the foundation of society,. are exceedingly remarkable. Not only is
die mother the head and source and only bond of union of the family; in the
most primitive part of the hills, the Synteng country, she is the only owner
of real property, and through her alone is inheritance transmitted. The
father has no kinship with his children, who belong to 'their mother's dan.
What he earns goes to his own matriarchal stock, and at his death his bones
are deposited in the cromlech of his mother's kin. In Jowai he neither lives
nor eats in his wife's house, but visits it only after dark. In the veneration of
ancestors, which is the foundation of tribal piety, the primal ancestress and
her brother are the only persons regarded. The flat memorial stones set up to
perpetuate the memory of the dead are called after the woman who repre-
sents the dan, and the standing stones ranged behind them are dedicated
to the male kinsmen on the mother's side. In harmony with this scheme of
ancestor-worship the other spirits to whom propitiation is offered are mainly
female, though here male personages also figure. The powers of sickness and
death are all female, and these are the most frequently worshipped. The two
protectors of the household are goddesses, though with them is also revered
the first father of the dan. Priestesses assist at all sacrifices, and the male
officiants are only their deputies. In one important state, Khyrim, the high-
priestess and actual head of the state is a woman, who combines in her
person sacerdotal and regal functions. 8
The centre of Khasi life is the village. It is usually situated
just below one of the hill-tops in which the country abounds.
Once built, it is never moved except under compulsion. It may
be destroyed by cyclones or marauders, but when the trouble
is over the inhabitants return and rebuild it on the old site.
Tne houses are closely packed, with no distinction between
those belonging to the chief's family and the remainder. All
around are the cromlechs and clan cemeteries, also the sacred
groves, dedicated to the village deity. These are taboo,- the
amber being reserved for the cult of the dead.®
The waste-land belongs to the village and is open to all for
7 Gurdon 6. 8 C. J. Lyall in Gurdon xix-xx. o Gurdon 33.
V MATRIARCHAL PEOPLES OF THE AEGEAN 153
thatching grass and firewood. The arable consists of dan
estates, owned collectively; sacerdotal estates for the upkeep
of the priests; and royal estates for the chief and h is family.
There are also a certain number of private estates acquired by
purchase. These offer the only exception, and that a limited one,
to the rule that the land belongs to the women. In the easterly
districts a man who has bought a plot of land is entitled to its
usufruct, but at death it reverts to his mother or her heiress.
In the west he has the same right, provided he is married, and
may even bequeath part of it to his children, but if he is
single it is simply counted as his earnings on behalf of his
dan. 10
The Khasis have a saying, 'From the woman sprang the dan*.
The dans are strictly exogamous. Marriage within the dan
is the greatest sin a Khasi can commit. He is excommunicated
and loses the right of burial in the dan sepulchre. Each dan
is divided into households. This unit, known as shi kpoh, 'one
womb’, comprises all those descended on the mother’s side
from a single ancestress down to the fourth generation. It is a
matriarchal oikos (pp. 109—10). The matcrfamilios administers the
cult of the family goddess, and also, if hers is the senior family,
that of the dan ancestress. The dan estate, from which a live-
lihood is guaranteed to all the dansfolk, is managed on behalf
of the senior materfamilias by her mother's brother. She is
succeeded by her elder sisters in order of juniority; in default
of sisters by her daughters, the youngest inheriting the house,
the elder only a share in the movables. Failing these, the
estate passes to the sisters’ daughters and then to the mother’s
sisters and their female descendants in the female line. 11
This does not leave much scope for the man. As a husband,
he is a stranger to his wife's people, who refer to him curtly
as a ‘begetter’. Marriage is monogamous to the extent that a
woman never has more than one husband at a time, but
divorce is so easy that, as Gurdon says, ‘the children are
10 ib. 82-7.
u lb. 77, 82-3, 88. The Khasi rule is the matriarchal counterpart of
Borough English, which is believed to have originated in conditions
of rapid expansion: Vinogradoff GM 314-5, F. Seebohm EVC 351,
Kovalevsky 135.
V
154 STUDIES IN ANCIENT GREEK SOCIETY
ignorant in many cases of their fathers' names'. This does not
worry them. They have been brought up in their mother's
house, and there, father or no father, they remain. 12
Public religion, as distinct from the clan cults, is administered
largely by male priests, but these are subject to a curious re-
striction. The priest performs the sacrifice, but a priestess
must always be present. The priest is her deputy. As Gurdon
points out, this is a survival from a time when the priestess had
officiated alone. 13
Where the chief is a man, his successors are his brothers,
beginning with the eldest, his sisters' sons, his sisters’
daughters' sons, and his mother’s sisters’ sons. In the absence
of male heirs the succession reverts to females — his sisters,
their daughters, and so forth. In Gurdon’s time the chief of
Khyrim was a woman, of whom he records the important
detail that she was in the habit of delegating her secular
duties to her son or sister’s son. 14 This suggests that the
chiefs, like the priests, have won their position by deputising
for women.
We can now see the whole history of succession in a new
light. In general, wherever the matrilineal rule has survived,
it takes the form of succession from mother’s brother to
sister’s son, which accordingly has come to be regarded as the
norm. Really it is transitional. The original form is preserved
in the Khasi clan, where succession passes from mother to
daughter, the men being excluded. This is modified .by de-
puting the woman's functions to the man — either the brother,
as among the Khasis and the Iroquois, or the husband, as in
the Roman monarchy (p. 97). The succession then passes
from man to man but in die female line — from mother’s
brother to sister’s son or from father-in-law to son-in-law.
And so we reach the patriarchal rule — the exact opposite of the
matriarchal — in which the succession passes from man to man
in the male line to the exclusion of the women.
# The Khasi matriarchate is unique in preserving as a func-
tional unity all those female rights which occur elsewhere
only in fragments or in traditions from the past. There is
ample evidence, however, especially in this part of Asia, that
12 Gurdon 81-2. 13 zb. 120-1. u zb. 70-1.
V
M AT,RI ARCH AL PEOPLES OF THE AEGEAN 1 55
institutions of this type were once general. The Garos of Assam
have the same rules of ownership and inheritance, with two
significant modifications. The husband enjoys the full usufruct
of his wife’s property, and the widow is required to marry her
Table VII
EVOLUTION OF PATRILINEAL SUCCESSION
M, man. W, woman. The inheritors are italicised.
1
M
1
W--
=M .
1
Direct matrilineal succession (mother to daughter)
l
r~i
1
M
W—M
r
M
1
W=
=M
Indirect matrilineal succession (mother's brother
to sister's son). The woman's rights are trans-
ferred to her brother.
r~r
1
M
W=M
1
M
W=
=M
Indirect matrilineal succession (lather-in-Iaw to
1
son-in-law). The woman's rights are transferred
to her husband.
m
1
M
W =M
1
M
W=
=M
Patrilineal succession (father to son).
r
M
W=M
youngest daughter’s husband, who thus inherits without in-
fringing the matrilineal principle . 15 In southern China there
are still tribes ruled by female chiefs . 18 and ancient China
was fully matriarchal. The women, according to Granet,
is Frazer TE 2. 323. 18 Briffault 3. 23.
156 STUDIES IN ANCIENT GREEK SOCIETY V
'transmitted their names to ‘their children,' the husbands were
only consorts annexed to a group of wives'. 1 ’ In the tenth
century a.d. northern Tibet was a highly organised matriarchate .
known to us from Chinese annalists, who called it Nu-kuo,
the Kingdom of Women. The queen’s husband was a nonentity
with no part in the government. This was in the hands of a
council of state convened by the queen and composed of the
palace women, whose decrees were executed by male officials
with the title of ‘women’s deputies ’. 18 Again we see how the
queen’s husband eventually became a king. This point is so
important that further illustrations are desirable. For these we
turn to Africa . 19
Among the Baganda the totem is to-day inherited from the
father, but formerly it was matrilineal. The old rule is still
preserved in the royal family, which being hieratic is naturally
conservative. The king is an absolute despot, yet strangely
dependent on two women. The queen and the queen-mother
both share the title of king. Each keeps her own court and
possesses her own estate administered by her own officials. One
of the queen-mother’s duties is to furnish the king with
daily gifts of food. Her death is regarded as a great calamity,
especially for him, and a successor is appointed from her clan
without delay, as though he could not survive without her.
The queen sits on the same throne as he does and takes the
same oath at coronation. She is chosen for him by the queen-
mother, and she is his sister.
Among the Baganda the queen-mother’s office is mainly
sacerdotal, but in the kingdom of Benin, in southern Nigeria,
besides holding her own court, she is consulted by the king on
all matters of state. She and her daughters live together. They
never marry but enjoy as many lovers as they please, drawn
from any rank of society. In Lunda the queen-mother reigns
jointly with the king. Her approval is required for all his acts,
her presence is indispensable at all his public appearances,
and her authority is supreme whenever he is absent.
All the more advanced African monarchies conform to this
17 Granet 343-4, cf. Bishop 305. is Briffault 3. 23-4, cf. 1. 647-53.
19 For the examples that follow see the authorities cited by Briffault
3. 28-36.
V
MATRIARCHAL PEOPLES OF THE yEGEAN 157
semi-matriarchal type, reflecting an antecedent stage in which
the king had been merely one of the queen’s husbands. In the
more backward kingdoms of Loango, Daura, and the Abrons
of the Ivory Coast, the king has hardly any power at all, and
he is' the son of a slave. In Agonna, Latuka, Ubemba, and
elsewhere, there is no king. The ruler is a queen, who does not
many but has servile lovers.
The patriarchal developments in these African kingdoms
have been fostered by wars of conquest, which, arising initially
from the process of tribal expansion, have been sharply in-
tensified by the repercussions of the slave trade and by
European and Mohammedan penetration. The primitive
matriarchate, founded on agrarian magic, has thus been
abruptly modified. We have indeed several instances of these
sturdy negresses leading their armies into battle against
European bayonets as energetically and hopelessly as our own
Boadicea, but the extension of warfare was bound in the long
run to weaken their authority. Their success in keeping such
a strong hold over their sons and husbands is due to their
sacerdotal functions, which being agricultural were the special
property of their-sex.
It must not of course be imagined that the powers of the
African, king are exclusively secular. On the contrary, he is
everywhere the high priest and in particular the supreme
rainmaker. Yet, as Briffault has shown, his sacral functions tell
the same tale. In Dahomey, where the king’s control of the
royal women is unchallenged, he is revered as a descendant of
the rain-god, who' is supposed to lodge with him in the palace;
yet it is his wives, under the title of the Mothers, who draw
the ceremonial water from the wells and perform the rain-
making magic. Hence we are not surprised to find that in
communities less -advanced than Dahomey the rainmakers are
regularly women. At Chigunda (Central Africa) the whole
tribe assembles for the rainmaking, but the actual ceremony is
conducted entirely by women. Among the Damaras prayers for
rain are offered by the chief’s daughter, who tends for the
purpose a sacred fire, which is never put out. So among the
Hereros, it is the chief’s daughter who prays for rain and
tends the sacred fire in the hut of his]; principal wife. The
V
158 STUDIES IN ANCIENT GREEK SOCIETY
reader will recall the ever-burning fire of the Roman Vestals,
who, as Frazer has shown, were originally wives of the Roman
kings . 20 \
Thanks to Frazer’s monumental researches it is now re-
cognised that the kingship is derived ultimately from agrarian
magic, its military and political functions being secondary.
The kin g secures his position by concentrating in his person or
under his control all the social energy directed towards the
fertilisation of man and nature. Endowed by this means with
the supreme power over his people’s welfare, he is revered as
a god and admitted to office by a special initiation — the rite
of coronation, which signifies that he has been born again, no
longer man but god . 21 The real nature of his exalted status is
vividly expressed in the words with which the Jukuns of
Nigeria acclaim a new king. They bow down before him and
cry: ‘Our rain, our crops, our health, our wealth !’ 22
If the king began as a mere consort of the royal women, it
becomes possible to understand what to modern minds is the
most puzzling of his primitive characteristics. It is again to
Frazer that we are indebted for the discovery that his tenure
of office was limited in early times to a prescribed period, at
the end of which he was put to death. When we consider the
marital customs of these African queens, who treat their
consorts as slaves, because they arc slaves, we can see that in
these conditions the king’s death was only an incident in a
women’s ritual cycle. Among the Shilluks of the Sudan, who
killed their kings within living memory, the princesses en-
joyed the same rights of free love, and in former days they
used to strangle the king with their own hands . 23
It was necessary for these ‘queens’ to conceive in order that
the earth might bear fruit. Their sexual life was a cycle of
mimetic magic. Accordingly, the procreator was imagined as
a god — in the first instance, no doubt, the god of the moon,
so Frazer GB-MA 2. 228. 21 Hocart 70-98.
-2 Meek 137* Not only was die Jukun king liable to be put to death if he
failed to produce plenty, but, though his authority was nominally absolute,
he was so hedged round with taboos that the real power rested with the
priests who acted as intermediaries between him and the people: Meek
333 - 4 *
23 Briffault 3. 36-7, Frazer GB-DG 17-8.
V MATRIARCHAL PEOPLES OF THE A 3 GEAN 159
which in primitive thought is the cause of pregnancy in
women and fertility in the soil; and after serving their purpose
the men in whom this god was embodied were put to death.
They had to die in order that the crops might live; This
ritual, which inspired the myths of Ishtar and Tammuz, Isis
and Osiris, Venus and Adonis, is the precursor of the Greek
sacred marriage, in which it was adapted to the conditions of
monogamy.
No one can study these Bantu monarchies without recalling
the kingdom of the Pharaohs. In ancient Egypt royalty was
transmitted in the female line . 24 The children of a royal
mother were royal, but the king could only secure his status
for his sons by marrying one of his sisters or a daughter of his
mother’s sisters . 28 This is the rule of matriarchal endogamy,
observed in ancient Egypt as among the Baganda to-day.
If the king’s mother was royal, he reigned in his own right,
while she occupied the same exalted position as the Bantu
queen-mother. The two are sometimes represented on monu-
ments as seated side by side . 26 If he was not of royal birth, he
reigned by right of marriage. 2 ’ Just as he was the god incarnate,
so the queen was a ‘wife of the god’ with a status hardly inferior
to his own. The celebrated Hatshepsut of the XVnith
Dynasty ruled the country for over thirty years in partnership
first with her father and later with her nephew, Thothmes
III, who, great man though he afterwards became, played
second fiddle till she died. Tutankhamen was overshadowed
in the same way by the queen Ankhsenpaaten and the queen-
mother Neffetiti, the energetic wife of Amenhetep IV . 28
If the king married outside the royal house, the succession
reverted to die female line, and consequently the founder of
a new dynasty usually took the precaution of marrying into the
old . 29 This practice persisted down to the days of Antony and
Cleopatra, when the Egyptian monarchy at last came to an
24 Petrie HE 2. 183, cf. SLAE 110-1.
28 Petrie HE 2 . 95-6. Cf. Koschaker 81.
26 Petrie HE 1. 114.
27 ll. 2. 240, Budge HE 4. 145.
•28 Hall AHNE 232, 308.
29 Revillout 2. 57, H. R. Hall in CAH 1. 279, cf, Bancroft 2. 142;
D.S. 17. 107, Arr. An. 3. 22, 5.
l6o STUDIES IN ANCIENT GREEK SOCIETY V'
end. It is the same principle as we have traced in the Sabine
and Etruscan dynasties at Rome.
Where ancient Egypt differed from the Bantu kingdoms, as
they are to-day, is that in Egypt the whole -of society was more-
or less matriarchal. The normal rule of inheritance was that
a man's property passed to his eldest daughter, though he
might bequeath specific goods to his sons . 30 If the woman
owned the property in the second generation, how, we ask, had
the man come to own it in the first? The answer is that strictly
speaking he did not own it at all. He merely enjoyed its use
by right of marriage. And this takes us back to the Garo
modification of the Khasi rule, which vests all property rights
in the women. Accordingly, following the Pharaoh's example,
the son married his sister. The Egyptian brother-and-sister
marriage was dictated by the assertion of male property rights
within a matriarchal system. As Petrie has put it, ‘sister-
marriage reconciled matriarchal property with paternal in-
heritance ’. 31
Under the Old Kingdom the status of women had been high
enough to qualify them for administrative fuhctions in the
public service, such as local prefectures, and the wife’s position
in the family was at least equal to the husband’s. But, begin-
ning with the Xllth Dynasty, we meet signs of a change, which
become still more pronounced towards the close of the Middle
Kingdom, promoted perhaps by the Hyksos kings, who were
pastoral nomads from the north. We now find that besides the
principal wife, the nept pa or mistress of the household, noble-
men were permitted to marry a ‘wife of the second degree '. 32
The priesthoods too fall increasingly under male control, and
women withdraw from public life. The manner of their with-
drawal is characteristic. ‘The inscriptions of Beni Hassan’,
according to Revillout, ‘prove that in this period, where
governmental functions were hereditary, subject to the
approval or veto of the sovereign, the woman transferred her
3 ° Breasted 86, H. R. Hall in CAH i. 279.
31 Petrie SLAE no.
3 - Revillout 2. 31, 39, 57-8. The institution of the s'econd wife is
recognised in Babylonian and Hittife law (Cuq 471) and is found also
among the ancient Irish (Dillon 38).
V MATRIARCHAL PEOPLES OF THE A 5 GEAN l6l
rights in them to her son or husband .’ 83 The woman was
supplanted by the ‘woman’s deputy'.
‘In Egypt,’ according to Hall, ‘there were always strong traces
of Mutterrecht , but none in Babylonia.’ 3 '* I hesitate to challenge
his authority. Robertson Smith believed that all the Semitic
peoples were originally matriarchal , 3 3 and matriarchy seems
to have left something more than a trace in the early Sumerian
city-states , which Hall himself described as follows:
Each city was ruled by a hereditary governor, who was also high-priest of
the local god and bore the tide pat/si, which signified that its possessor was
the earthly vice-gerent of the god s. The Sumerian language possessed a word
denoting the ruler of a higher political organisation: this was lugal, ‘king'
(literally 'great man’). This word had no theocratic connotation and ... it
seems to have been assumed by any patesi who succeeded by force or fraud
in uniting several cities under his government. 38
The office of patesi was theocratic, that of lugal rested on
military power. This distinction is in keeping with the normal
development of the kingship in the decline of the matriarchate.
At the beginning of Sumerian history we find Baranamtarra,
wife of Lugalanda, the patesi of Lagash, ruling the city jointly
with her husband. She bears the honorific title of ‘the Woman’,
and she keeps her own court, the ‘House of the Woman', as
distinct from the ‘House of the Man', which belongs to the
patesi. The wife of the next patesi, Urukagina, enjoyed a
similar status. Her name was Shagshag, her title ‘the goddess
Bau’. The chief minister of state was styled, under Lugalanda,
‘scribe of the House of the Woman’, and, under Urukagina,
‘scribe of the goddess Bau'. He belonged therefore in both
reigns, to the patesi’ s wife’s retinue. In both reigns, moreover,
official documents were dated in her name. All this suggests,
as Langdon has remarked, that the patesis were merely consorts,
the real authority being vested in their wives . 37 If this is not
matriarchy, it is very like it. Nor were such conditions peculiar to
Lagash. At Zabshali, and again at Anshan, we hear of a patesi
married to a daughter of a lugal; and in at least one instance,
at Markhashi, a lugal’ s daughter actually held office as patesi . 36
33 Revillout2. 57,91. 84 HaUAHNE205. 33 Robertson Smith KME A.
38 Hall AHNE 178-9. 37 S. H. Langdon in CAH 1. 385-6.
38 R. C. Thompson in CAH 1. 509-10.
L
V
l6z STUDIES IN ANCIENT GREEK SOCIETY
Matriarchal institutions have been traced in ancient Elam, 3 °
and there they were transmitted to the Persian Emperors.
Readers of ABschylus wilf remember the majestic figure of the
Queen-Mother Atossa, who during her son’s absence ruled the
kingdom . 40 Dareios, the father of Xerxes, was her second
husband. Her first was her brother, Kambyses, and after his
death, according to Herodotus, she continued to ‘hold all the
power ’. 41 That no doubt is why Dareios married her. A later
Dareios, contemporary with Alexander the Great, succeeded to
the throne by marrying one of his sisters, who were all ‘prin-
cesses of the blood royal ’. 42 The strenuous part played by
women in the dynastic struggles of the Macedonian monarchy
suggests that it too may have contained matriarchal elements . 43
Be that as it may, brother-and-sister marriage is definitely-
attested for Alexander’s successors, the Ptolemies, Arsacids,
and Seleucids. The Ptolemies took it over from the Pharaohs,
the Arsacids and Seleucids from the Persians. Thus, Laodike,
daughter of Antiochos III, was married in turn to her three
brothers, Antiochos, Seleukos IV, and Antiochos IV. By
Seleukos she had a son, also named Antiochos, who was pro-
claimed king in boyhood under the regency of her third
brother, Antiochos IV, who then married her. Tarn says that
the regent’s motive for the marriage was to secure the suc-
cession for his ward . 44 But that was already guaranteed by his
parentage. It is much more likely that he wanted it for himself.
And he got it. Shortly afterwards the boy was assassinated, and
Antiochos IV became king. Who was the murderer?
Arrian says that Asia Minor had been ‘ruled by women’
ever since the legendary days of Semiramis . 46 This may be.
an exaggeration, though, as a native of the country, he ought
to have known. There are two reasons why in modern histories
of the Near East the status of women has been neglected. One
30 Konig MTAE.
40 A. Per. 153-60. ./Eschylus seems to have been well acquainted with
Persian life: Konig RI 88-90.
41 Hdt. 7. 3. 4, cf. 3. 31, 68, 88.
42 Arr. An. 2. 11-2, cf. Luc. Satr. 13. 5. The Persian Magi practised
matriarchal endogamy: Xanth. 28.
43 Plu. Alex. 9.
44 Tarn 185. See further Wesendonk VA. 40 Arr. Ini. 1. 23. 7.
V
MATRIARCHAL PEOPLES OF THE AEGEAN 163
is the general lack of understanding of what matriarchy is
and the rather indolent assumption that ancient society must
have resembled our own except in so far as it is definitely proved
to have been different. The other lies in the ancient documents,
which, beingmainly concerned with political life, give a one-sided
picture. After the decline of the matriarchate the women’s
publicly acknowledged privileges were confined to religion,
but this did not prevent them from exercising an unobtrusive
influence on secular affairs. Long after surrendering the form
they retained the reality, and so developed one of the charac-
teristics of their sex.
2 . The Lycians
Let us now draw the net closer. Linguistic evidence has led
us to conclude that the Greek-speaking immigrants into the
iEgean came trader matriarchal influences (p. 146). What
support is there for this conclusion in their traditions about
themselves?
The Algean basin was never completely hellenised. In the
north it remained exposed to fresh irruptions — Thracians,
Phrygians, and later Macedonians, Gauls, and Slavs. In
Anatolia, it was only after the conquests of Alexander that
Greek speech penetrated into the interior. Behind Aiolis lay
the Phrygians, behind Ionia the Lydians, behind the Dorian
settlements further south the Carians and Lycians. A non-
Greek language was still spoken in parts of Crete as late as the
fourth century B.c. 46
The Lycians were so called because their national^ god,
Apollon Lykios, was worshipped as a wolf ( lykos ). 47 His
mother, Leto, is said to have been changed into a wolf before
his birth, or led by wolves to the spot where he was bom. 48
Their own name for themselves was Trmmli, vocalised in
46 Nilsson HM 65-6. 47 Scr. Myth. Gr. 77.
48 Ail. NA. 10. 26, Ant, Lib. 35. Apollo appears with attendant wolves
on coins of Tarsos: Imhoof-Blumer 171. This seems enough to fix the
meaning of 11 . 4. xoi which is simply epic for AwoyeveT, cf. 11 .
2. 54 TTuAoiysvIos, 3. 1 82 iioipriyEvfc, and see W. G. Headlam in G. Thomson
AO 2. 10. There is consequently no need for Kretschmer’s hypothetical
*A 0 *n (Hittite Itg|fl)=AuKlrj (SLS 102). See further his NLKV 14-7*
164 STUDIES IN ANCIENT GREEK SOCIETY V
Greek as Termilai. From Egyptian annals, in which they
appear under their Greek name (Luka), we learn that in 1248
b.c., together with other ABgean peoples, they had been raiding
the Nile Delta. A century later a section of them and the
Carians migrated through Pamphylia and
Cilicia into northern Palestine, where they
became known as the Philistines. 49
Reference has already been made to
their matriarchal institutions (pp. 99,
142). Descent was matrilineal. When
Plutarch mentions the Lycian clan Ioxidai
(p. 122), he calls them ‘Ioxidai or Ioxides’,
implying that the feminine form was the
proper one. 80 Succession too was matrilineal.
Daughters inherited in preference to sons. 81
The basic unit of society, attested by sepulchral
inscriptions, was the matriarchal household.
Some of these inscriptions contain a formula of the familiar
matriarchal type: ‘Neiketes son of Parthena. . . . Neiketes son
of Lalla. . . . Eutyches, father unknown. . . . Alexandras,
father unknown.’ 62 Systematic excavation in this area will
add much to our knowledge of the Anatolian matriarchate.
In Greek tradition, it was with Lycian aid that King Proitos
occupied and fortified the stronghold of Tiryns, 68 one of the
most important Mycenean sites in the Argive plain. In the
same generation Bellerophon, son of Glaukos and grandson of
Sisyphos, after sojourning at the court of Proitos, migrated to
Lycia, where he married the king’s daughter and received a
share of the kingdom. He had a daughter Laodameia and a son
Hippolochos. Laodameia became by Zeus the mother of
Sarpedon, who led the Lycians to the Trojan War. Hippolochos
fig. 6 . A Philistine:
Egyptian painting
40 H. R. Hall in CAH 2. 282-4. eo Plu. Ehes. 8. 61 Nic. Dam. 129.
62 TAM. 2. 176. a. 48, b. 20, 46 TTOTp6s dS^Xou, cf. 2. 601. It is
possible that in these cases the mother was a priestess of the same type as
the Babylonian Nin-An, ‘bride of God/ ' Sargon, whose mother was pro-
bably a Nin-An, ‘knew not his father': R. C. Thompson in CAH j. 536-7.
63 Apld. 2. 2. I , where Bellerophon's father-in-law is given as Amphianax
or Iobatcs. The latter was a Lycian name: TAM 2. 283. According to
II. 6. 170 sch. he was Amisodaros, cf. ll. 16. 328. Amisos was a town in
Paphlagonia (Str. 68-71) and for the termination cf. Pixodaros (p. 167).
V
MATRIARCHAL PEOPLES OF THE jEGEAN 165
was the father of another Glaukos, Sarpedon’ s companion at
Troy . 84 When the Greeks colonised Ionia, members of this
family were chosen as kings at Miletos and elsewhere . 5 8
Another branch remained in Lycia at Xanthos, where there
was a townland called Glaukou Demos . 88
Table VIII
DESCENDANTS OF SISYPHOS
Sisyphos
Glaukos Amisodaros
Bellerophon = Philonoe
Hippolochos Laodameia— Zeus
I I
Glaukos Sarpedon
The fact that the Lycian leader at Troy was Sarpedon and not
Glaukos attracted the attention of the ancient Homeric com-
mentators, who explained it quite correctly as a mark of
honour for his mother.” Since Bellerophon had attained royal
rank by marrying the king's daughter, the succession passed
through her daughter. This, as we have seen, is a form of
indirect matrilineal succession.
The Glaukidai must have been Greek-speaking, otherwise
the Ionians would not have chosen them to be their kings.
They cannot have learnt Greek in Lycia, where the native
language survived into the Christian era, and therefore the
stock of Sisyphos must have been Greek-speaking when they
left the Peloponnese. This is just what our linguistic analysis
has led us to expect. A Greek clan, settling among an alien
Mil. 6. 152-206. There is an Irish tradition very similar to the story
of Bellerophon: Dillon 35. Traces of succession from father-in-law to
son-in-law are found in Scandinavian mythology: Chadwick OEP 312.
88 Hdt. 1. 147.
88 Alex. Polyh. 82-3.
67 Eust. ad II. iz. 101. Similarly, the Kinyradai of Cyprus, descended
from Teukros and a daughter of Kinyras, owed their priesthood of Aphrodite
to the latter (Paus. 1.3.x, Tac. H. 2. 3.): see p. 513.
V
l 66 STUDIES IN ANCIENT GREEK SOCIETY
matriarchal people, rises to power by conforming to the
indigenous rule of succession.
This is not the only case of its kind. The Dorians of Argolis
were organised in their three ancestral tribes (p. 102), but
besides these they had a fourth, the Hyrnatheis, drawn from
the conquered population. 68 The story of their eponym
Hyrnetho (Doric Hyrnatho) was as follows. Temenos, the
Dorian chief to whom Argolis had been assigned, offended his
sons by favouring Deiphontes, who had married his daughter
Hyrnetho. Fearing to lose the succession, his sons suborned
some criminals, who waylaid him and killed him, but the old
man lived long enough to bequeath the kingdom with his
dying breath to his daughter and son-in-law, who, after being
confirmed in possession by the people, reigned jointly. 69 The
story may not be historical, but that does not affect its value
as evidence of early custom. Besides illustrating the conflicts
that accompanied the transition from mother-right to father-
right, it gives us a Greek instance of the principle that, where a
man succeeds his father-in-law as king, he does so as consort of
his queen, who reigns in her own right. 60
3. The Carians and Leleges
The Carians and Leleges both belonged to the Anatolian
seaboard, and the distinction between them is somewhat in-
definite. Herodotus regards the Leleges as a branch of the
Carians that retained the old national name. Other views were
that they were a distinct people reduced by the Carians to
serfdom, and that originally they had been confined to Samos
and Chios. 61 In historical times they were little more than a
68 IG. 4. 517, St. B. Agaves, SIG. 594 n. 4.
69 Nic. Dam. 38, Apld. 2. 8. 5.
60 So at Megara: Sikyon married a daughter of Pandion and claimed
the succession against his brother-in-law Nisos; it was divided between
them; Nisos was succeeded by his son-in-law Alkathoos and he by his
son-in-law Telamon: Paus. 1. 39. 6, 41. 6, 42. 4. At Corinth Jason suc-
ceeded by marriage to Medea (Paus. 2. 3. 10); Oros of Troizen was suc-
ceeded by his daughter’s son (Paus. 2. 30. 5), cf. 4. 30. 3 and see farther
7. x. 3* 8. 5. 6, Parth. 1, D.S. 4. 33.
61 Hdt. 1. 171. 2, Phil. Theang. 1 =FHG. 4. 475, Pher. in, Str. 321, 661.
V
MATRIARCHAL PEOPLES OF THE JE GEAN l6 J
memory, whereas the Carians were universally familiar as the
non-Greek inhabitants of the country that bore their name.
The principal Greek settlement in Caria was Halikarnassos,
the birthplace of Herodotus. The historian himself was pro-
bably of Carian extraction, for the names of his father and
uncle, Lyxes and Panyasis, are not Greek.® 8 Though more
exposed to Greek influence than the Lycians, they too pre-
served their language and culture. Herodotus must have known
them well, and, since he describes the Lycian matriarchate as
unique (p. 144), it would seem to follow that the Carians of
his day were patrilineal. Even here, however, it is necessary to
make reservations.
The best-known of the Carian kings was Mausolos, who
reigned in the fourth century. His wife was his sister Artemisia.
He had two brothers, Idrieus and Pixodaros. Idrieus was
married to another sister, Ada. Mausolos died childless and was
succeeded by Artemisia, who erected to his memory the
famous Mausoleum. She was succeeded by Idrieus, and he by
Ada. This lady was expelled by Pixodaros, who submitted to
the Persians and left the kingdom to the Persian satrap, who
married his daughter. Finally the satrap was expelled by
Alexander the Great at the instance of Ada, who thus reigned
once more in her own right.® 3 A hundred years after Herodotus
we find 'the Carian dynasty observing the same rule of matri-
archal endogamy as the Pharaohs.
We learn from Herodotus himself that at the time of the
Persian War his native Halikarnassos was under a Carian
queen, who, to judge by her name, Artemisia, belonged to the
same dynasty. Her mother was a Cretan; her father was named
Lygdamis. Her husband was dead, but, though she had a
grown-up son, she retained the royal power ‘out of sheer
manly spirit*. Her domain extended to the adjacent islands of
Kos, Kalymnos, and Nisyros. 64 When Xerxes invaded Greece,
® 2 Suid. 'HpiBoros, Dur. 57. The survival of such place-names as
Ouassos and Onzossyasos (SIG. 46) suggests that Carian continued to be
spoken in Halikarnassos itself.
® 3 Str. 656-7, Arr. An, 1. 23. 7-8. For Pixodaros cf. SIG. 169. 16 and
see p. 164 n. 53. Arrian l.c. says that brother-and-sister marriage was a
Carian custom.
® 4 Hdt. 7. 99.
PREHISTORIC PEOPLES OF THE AEGEAN Map III
V MATRIARCHAL PEOPLES OF THE AEGEAN 169
she furnished, him with a contingent of five warships, com-
manded by herself. At the Battle of Salamis, when the Persian
rout had begun, her flagship was hotly pursued by the
Athenians, but she saved herself by adroitly turning about and
ramming a Persian vessel. The Athenians took this to mean
that she was deserting to their side, and gave up the chase.
The Persians, on their part, supposed that the ship she
rammed must be an enemy; so Xerxes, watching the battle
from the shore, and disgusted with the failure of his own
admirals, made the famous remark, ‘My men have become
women and my women men'. 66 The special interest of this in-
cident is that on board one of the Athenian ships, perhaps an
eye-witness, was the dramatist whose greatest character sur-
passed even Artemisia in masculine strength of purpose.
The Ionian conquerors of Miletos took Carian wives, who,
resenting the slaughter of their menfolk, refused to eat with
their new husbands or call them by their names. 60 This implies
that in the early days of the colony the women had maintained
to some extent their native organisation. At Teos, another
Ionian settlement, there has been recovered a list of annual
magistrates. 67 In each case the man's name is followed by those
of his clan and pyrgos. The pyrgos was his village, equivalent to
the Attic deme. And in 1 1 cases out of 25 the clan and village
have the same name, e.g. ‘Euthyrrhemon Boides of Boios*.
This means that the identity of the two units was still largely
intact. The clan names themselves are significant. One of
them, Philaides, is Attic (p. 121); another, Kothides, comes
from Euboia; 68 a third, Maliades, from Thessaly. 69 Several,
such as Bryskides and Daddeios, are Carian. 70 Since these
Carian clans remained in occupation of their native settle-
ments under their native names, they must have preserved
their native institutions; and, if this happened at Teos, it must
have happened in other Ionian colonies.
In prehistoric times the Carians and Leleges had extended
far beyond Caria. They are said to have been driven from the
66 Hdt. 8. 87-8, cf. Ar. Lys. 675.
60 Hdt, 1. 146. 3. Herodotus says the Ionian women’s costume was
of Carian origin (5. 88).
67 CIG . 3064. 68 Str. 447. 60 Cf. Str. 633. 70 Cf. Paus. 7. 3. 6.
V
170 STUDIES IN ANCIENT GREEK SOCIETY
Troad after the Trojan War .’ 1 The old name of Kos was
Karis, and there was a townland in Chios called Karides .’ 2 They
are mentioned as early inhabitants of Naxos ,’ 3 where we meet
with the personal name Lygdamis , 74 and Naxos itself seems
to be connected with the Carian town of Naxia .’ 6 The Carians
of Naxos are said to have come from Lamia in the extreme
south of Thessaly . 76 Epidauros and Troizen, on the Argive
coast, were Carian settlements.” The acropolis of Megara
was called Karia after King Kar, the ‘Carian *. 78 The cult of
Zeus Karios, centred at Mylasa, the Carian capital, is found
in Boeotia 79 and again in Attica . 80
Another early lung of Megara was Lelex, and Leleges from
Megara were die original founders of the Messenian Pylos . 81
Lelex was also the first king of Sparta, whose earliest inhabi-
tants are described as Leleges . 82 We also hear of Leleges in
Leukas, Akarnania, Lokris, and Boeotia . 83 Lastly, Thucydides
says that the Carians were expelled from the Cyclades during
the Minoan thalassocracy, and he adds that in his own life-
time, when some ancient graves were dug up in Delos, more
than half the corpses were identified by their accoutrements as
Carian . 84
Nevertheless, the Carian domain has definite limits. It is
71 Str. 321, cf. II. 21. 85-8. Antandros, Skepsis, Pedasos, Gargara, Assos
had all belonged to the Leleges: Str. 605-10.
72 Hell. 103. Eph. 34. The early inhabitants of Samos and Chios are
described as Carians: Pans. 7. 4. 8-9, Str. 637.
73 D.S. 5. 51. 74 Hdt. 1. 61. 4, cf. 7. 99. 2. 76 Alex. Polyh. 54-5.
78 D.S. 5. 51. 77 Arist. fr. 49i=Str. 374. 78 Paus. 1. 40. 6.
79 Hdt. 1. 171. 6, Str. 659, Phot. K&pios Zoft. 80 Hdt. 5. 66.
81 Paus. x. 39. 6, 4. 36. x. Another Lelex setdement, identified by
its name, was Pedasos in S. Messenia (II 9. 152). Expelled from the
Trojan Pedasos (see n. 71) the Leleges fled to Halikarnassos, where they
founded Pedasa (Str. 61 1). The Messenian Kardamyle (II. 9. 150) was pre-
sumably founded by Leleges from Chios, where there was a town .of the
same name (Th. 8. 24. 3).
82 Paus. 3. x. x, cf. 3. 12. 5. A son of this Lelex founded Andania (Paus.
4. 1. 2) and his daughter gave her name to Therapne (Paus. 3. 19. 9).
The Spartan Leleges were agricultural: their king, Lelex, had a son Myles,
the miller/ who ground corn at Alesiai (‘grinding women’), and a grandson
Eurotas, who drained the Eurotas valley: E. Or. 626 sch.
83 Arist. fr. 56o=Str. 321-2; see further pp. 425-30.
84 Th. 1. 4. 8.
V MATRIARCHAL PEOPLES OF THE jEGEAN 171
bounded by a line drawn from Leukas to Lamia and thence
across to Chios. North of this line the prehistoric inhabitants
remembered by the Greeks were Pelasgoi.
4. The Pelasgoi
The Pelasgoi survived, still speaking their own language, at
several places in the north iEgean — Akte on the Macedonian
coast, Kreston somewhere in the same region, Lemnos and
Imbros, 86 and Plakia and Skylake in the territory of Kyzikos
on the Propontis. 86 They are also recorded in Samothraike, the
Troad, Lydia, Lesbos, and Chios. 87
In Greece proper they left their name in the ancient shrine
of Zeus Pelasgios at Dodona, 88 and in the Thessalian plain,
which was known as Pelasgikon Argos or Pelasgiotis. 80 They
are mentioned as early inhabitants of Bceotia and the Pelopon-
nesian Achaia, 90 and more especially as the aboriginal popula-
tion of Attica, Argolis, and Arcadia. 91 Near Olympia there
were remnants of a tribe called the Kaukones, who had once
ranged over the whole of Elis. These too were probably
Pelasgoi. 92 A tribe of the same name is mentioned along with
Pelasgoi in the Iliad as allies of the Trojans, and the name re-
appears further north in the Kaukones or Kaukoniatai of
Paphlagonia on the Black Sea coast. 93 There is no trace of
88 Th. 4. 109. 4 (cf. Str. 331. 35). Hdt. 1. 57, 5. 26 (cf. Str. 221),
4. 145. 2. The traditions relating to the Pelasgoi were collected by Hellanikos
of Lesbos, whose Ploronis was probably based on an epic with that title:
Pearson 159.
88 Hdt. 1. 57, cf. Deioch. $- 6 =FHG. 2. 17-8, Hec. 205, Eph. 104.
87 Hdt. 7. 42, Str. 221, 621, St. B. Niv6ii.
88 II. 16. 233, Str. 327, Plu. Pyrrl. 1.
89 'll. 2. 681, 840, Str. 221. 443. The word fipyos, which meant 'plain'
(Str. 372, cf. J. D. Denniston ad E. Io 1), was probably Pelasgian; and if, as
will be argued later (p. 396), the Achsan name for Thessaly was Hellas, then
the Homeric '0A<&6cc Kctl plow 'Apyos (Od. 1. 344) may be interpreted as a
description of that region by its two alternative names.
99 Str. 410, Hdt. 7. 94.
91 Hdt. 1. 57, 4. 145. 2, 6. 137. I, cf. Th. 2. 17; E. fr. 228, Hdt. 2.
171; Hdt. 1. 146, Paus. 8. 1. 4.
92 Str. 345, 542, Od. 3. 366.
93 II. 10. 429, 20. 329, Str. 345. The Kaukoniatai of Str. 345 are evidently
the Kaukones of Str. 541-2.
V
172 STUDIES IN ANCIENT GREEK SOCIETY
Pelasgoi in the southern Peloponnese or the Cyclades, but they
are mentioned in the Odyssey as one of several peoples inhabiting
Crete . 04
The name, according to Kretschmer, is an ethnical derivative
of pelagos . 06 This is an Indo-European word for a level surface, a
plain, but in Greek it was applied to the sea (cf. Latin aequor).
The current Greek for sea’ was thdlassa, which is not Indo-
European. Was this borrowed by the Greek invaders of the
TEgean from the ‘people of the sea’ they found there — the
Pelasgoi?
Though widely scattered, their culture seems to have been
homogeneous. One of their distinctive place-names; Larisa, is
found in several parts of Thessaly, Attica, Argolis, Elis, Crete,
the Troad, Aiolis, and Lydia . 06 The worship of Hephaistos,
the fire-god, which was certainly pre-Hellenic, was centred at
Athens and Lemnos . 07 He also figures in the Pelasgian cult
of the Kabeiroi, which survived in Samothraike, Lemnos, and
Imbros . 08 It has already been argued by A. B. Cook that
Hephaistos was a Pelasgian divinity . 00 So in all probability
was Hermes. He too was associated with the Kabeiroi, and
he had a non-Greek cult in Imbros . 100 His oldest seats on the
mainland were in Arcadia and Attica. He is said to have been
born on the slopes of Mount Kyllene in Arcadia, where he was
04 01 19. 177.
08 Kretschmer GGD 16-7, but cf. Cuny 21. In default of independent
evidence all such etymologies must be treated with reserve.
06 1 /. 2. 841, Str. 430, 440, 620-1, Paus. 2. 24. x, 7. 17. 5. The Cretan
Larisa was absorbed in the later Hierapytna (Str. 440). Larisa was a daughter
of Pelasgos (Paus. 2. 24, 1) and Larisa Kremaste was also known as Larisa
Pelasgia (Str. 435).
07 II 1. 593, Philoch. 6, S. Ph. 986-7, Dion. Chalc. z=IHG. 4. 393,
II. 2. 722 sch. V, Nic. Th. 472 si., Lyc. 224 sch. ' Hephaistos figures
on the coinage of Kyzikos, Bithynia, and Lydia: Famell CGS 5. 394.
08 Hdt. 2. 51, 3. 37, Str. 472, Paus. 9. 25. 5-10.
00 Cook Z 3. 226; see also K. Bapp in Roscher LGRM 3. 3040-1.
100 St. B. "ipppos: he was known there as Imbramos. Hermes appears
on Imbrian coins (Head 261), also in Lemnos (A. A. 295-6) and Thrace,
whose kings claimed descent from him (Hdt. 5. 7, cf. Famell CGS 5.
77 ). no doubt the Thracians had taken him over from the Pelasgoi. On his
name sec Kretschmer NKLV 3-4.
V
MATRIARCHAL PEOPLES OF THE AEGEAN 173
worshipped as an ancestor god, 101 and at Kyllene in Elis his
image consisted simply of a penis erectus , loa analogous to the
phallic effigies called hermat, whose origin was ascribed to the
Pelasgoi. 103 The Eleusinian Hermes, clan ancestor of the
Ketykes, was connected with the myth of Daeira (p. 128) and
that in turn with the Samothracian Mysteries. 1 ® 4
Where had the Pelasgoi come from? Not from the south. In
Crete they are expressly distinguished from the Eteokretes or
True Cretans, 105 and they appear nowhere else in the southern
yEgean. Nor from south-western Anatolia. That belonged to
the Carians and Lycians. All the signs point to the north —
to the Macedonian coast, together with the islands of Samo-
thraike, Lemnos, and Imbros, which lie at the gates of the
Hellespont; and since we have traced them through the
Hellespont and Propontis along the north coast of Anatolia,
there is a strong case for placing their original home some-
where on the far side of the Black Sea.
Thucydides, who had ancestral connections with the north
coast of the ^Egean, describes the Pelasgoi of Akte, Lemnos,
and Attica as Tyrrhenoi (Tyrsenoi). 106 Sophoklcs applies the
same designation to the Pelasgoi of Argolis. 107 This was the
name by which the Greeks knew the Etruscans. According to
Greek tradition the Etruscans had migrated to Italy from
somewhere in the JEgean — Herodotus says, from Lydia; other
writers describe them as Pelasgoi from Thessaly, or from
Lemnos and Imbros. 108 Conversely, the Etruscans of Caere
claimed descent from Thessalian Pelasgoi. 100
101 Horn. H. 4. 1-7, Paus. 8. 17. 1-2, A. fr. 273. His nativity was also
located at Thebes and Tanagra (Paus. 8. 36. 10, 9. 20. 3) and he had an
important cult at Pheneos (Paus. 8. 14. 10, 8. 16. 1, cf. 8. 47. 4).
102 Paus 6. 26. 5.
103 Hdt. 2. 51, Paus. 4. 33. 3. The ithyphallic Hermes appears on coins
of Imbros: Head 261.
104 Paus. X. 38. 7, cf. Cic ND. 3. 22. 56, Prop. 2. 2. 9-12, Hdt. 2. 51;
Toepffcr AG 96, Lobeck 1215-1348.
105 Od. 19. 177.
100 Th. 4. 109. 4. The Cretan FtXxovfis and Etruscan Velchans (Lat.
Volcanus) may be assigned along with Hephaistos to the Tyrrhcnoi-Pelasgoi:
Kretschmer SLS 28. 109.
107 S. fr. 248.
108 Hdt. 1. 94, Str. 443, 221; Kretschmer EP. 100 Str. 220.
174 STUDIES IN ANCIENT GREEK SOCIETY . V
Tyrrhenos is an ethnical derivative of Tyrrha, a town in
Lydia. 110 The name was borne by a brother of Tarch'on, the
Greek form of Tarquinius. 111 Their father, Telephos, appears
in Italy as progenitor of the Tarquinii, in Lydia as king of
Teuthrania. 112 Lastly, some inscriptions discovered in Lemnos
are in a language closely related to Etruscan. Of the Lydian
language little is known, but enough to show that it belonged
to the same family. 113
Like the Etruscans, the Lydians practised pre-nuptial
promiscuity (p. 141) — a relic of group-marriage. The Etruscans
are known to have been matriarchal, and this makes it likely
that at the time of the migration the Lydians were matriarchal
too. We hear of three Lydian dynasties — the Atyadai, Herak-
leidai, and Mermnadai, the last being the house of Croesus.
The pedigrees are confused, but we learn that Sadyattes of the
Mermnadai married his sister, and that his son and heir,
Alyattes, did the same. 114 Herodotus says that in the preceding
dynasty the succession had passed from father to son, 116 im-
plying that it was patrilineal; but, while there is no reason to
doubt the fact, the implication is open to question. Brother-
and-sister marrikge also results in succession from father to
son, being designed for that purpose, but in origin it is
matrilineal: the son inherits properly from the mother. It is
possible therefore that the Heraldeidai followed the same rule
as the Mermnadai — in fact, more than possible, because there
are grounds for suspecting that the tradition given by Herodo-
tus has been tampered with. The founder of the dynasty, in
his account, was a son of Herakles by a Lydian slave girl, a
daughter of Iardanos. This is a striking deviation from the
version given by Sophokles' and others. Sold into slavery by
Eurystheus, Herakles was bought by Omphale, the daughter
of Iardanos, who was no slave girl but a queen, and since
her husband's death she had reigned alone. 116 The Lydian
110 EM TOppa; Toepffer AG 195. 111 Lyc. 1248.
112 Lyc. 1249 sch., St. B. Tapxciviov, D. H. AR. 1. 28. 1, D.S.4. 33. On
Telephos see Kretschmer NKLV 13-4.
113 Cortsen LID, Kretschmer SLS 28. 108, R. S. Conway in CAH 4. 408.
114 Nic. Dam. 63, Suid. 'AAvdrrns.
116 Hdt. 1. 7. 4. ns s. Tr. 252-3, D.S. 4. 31. 5-8, Hyg. F. 32.
V • MATRIARCHAL PEOPLES OF THE AiGEAN 175
Herakles in this story is the Etruscan Servius Tullius
(P- 97 ).
If the Lydians and Etruscans were matriarchal, so were their
kinsmen die Pelasgoi. The Pelasgoi of Lemnos figure in one
of the best-known Greek legends. After setting sail from Thes-
saly in quest of the Golden Fleece, the Argonauts put in at
Lemnos, which was then 'ruled by women* under Queen
Hypsipyle, daughter of Thoas. Some time before the Lemnian
women had given offence to Aphrodite, who afflicted diem
with a smell so unpleasant that their husbands deserted them.
The women replied by murdering their menfolk, all except
Hypsipyle, who spared her father. Jason, the captain of the
Argonauts, fell in love with her, and their son, Euneos,
founded the clan Euneidai (p. 122). 117
The meaning of this myth, first explained by Bachofen , 118 is
not open to doubt. It enshrines the memory of the Pelasgian
matriarchate, but in a degraded form, corresponding to the
subsequent degradation of the female sex:
Of all the crimes told in tales the Lemnian
Is chief, a sin cried throughout the world with such
Horror, that if men relate
Some monstrous outrage they call it Lemnian.
Abhorred of man, scorned of God,
Their seed is cast out, uprooted evermore;
For none respects what the gods abominate. 110
This reads like a curse on the old order by the new.
The Tyrrhenoi-Pelasgoi of Attica were a branch of the
Lemnian . 120 They had been employed by the Athenians to
build a wall round the Acropolis . 121 In those days there were no
slaves, and the freeborn Athenian boys and girls who went to
fetch water from the Nine Springs were constantly being
assaulted by the Pelasgoi, who accordingly were driven out of
Attica and settled in Lemnos . 182
Democratic Athenians were proud of their Pelasgian origin.
They called themselves ‘sons of the soil '. 188 Herodotus describes
117 Hdt. 6. 138, A.R. 1. 609-23, Apld. 1. 9. 17, Hyg. F. 15.
118 Bachofen 84-7. 118 A.C. 631-4.
120 Th. 4. 109. 121 Hdt..6. 137. 2, D.H. AR. 1. 28. 4.
182 Hdt. 6. 137. 3-4. 188 Ar. V. 1076, E. Jo 20.
V
176 STUDIES IN ANCIENT GREEK SOCIETY
them as hellenised Pelasgoi. 124 One of their early kings was
Kekrops, 125 the founder of matrimony (p. 142). Before his
time the women had mated indiscriminately and named their
children after themselves. This is exactly what we are told of
the Etruscans (p. 142). x
The Etruscans are further connected with Anatolia — and
not with Lydia only but with Caria and Lycia — by numerous
parallels in place-names. Moreover, throughout the ABgean
basin and the Anatolian hinterland as far as Cilicia in the
south and the Caucasus in the north we encounter place-names
based on certain non-Indo-European elements (- nth - nd~,
-ss- t -tt-), e.g. Korinthos, Kelenderis, Myndos, Parnassos, Knos-
sos, Hymettos, Adramyttion. 126 The word thdlassa (Attic
thdlatta ) belongs to the same type. They are naturally most
plentiful in Caria and Lycia, where the pre-Hellenic languages
lasted longest, but their wider range shows that the -Egean
basin must once have constituted a uniform linguistic domain
extended from Anatolia.
Lastly, the speech of the Etruscans was related to languages
still spoken in the Caucasus. This discovery was made fifty
years ago by Thomsen, and has been confirmed by Marr. 122
That is as far as I can go. The problems raised by the
Caucasian affinities of Etruscan and other Asianic languages
have been complicated and extended by the discovery of a
common linguistic substratum covering the whole region from
the Black Sea to Syria and from the Aegean to Sumer. 128
Further, if these languages came from South Russia, where the
Indo-European diaspora is believed to have taken place, some
of the non-Indo-European elements in Greek, which are very
deep-seated, may be as old as Greek itself. The very concept
124 Hdt. 1. 57. 3. las Hdt. 8. 44. 2.
120 Kretschmer EGGS 401—6, ASK 92—6, Schwyzer 1. 60—1, Eisler SAQ,
Blegen CG, Haley CG, Nilsson HM 64—5. The forms in -nth—, common
in Greece, do not occur in Anatolia, with the exception of Xanthos (Lycia,
Troad), and conversely there appear to be only four instances of -nd-
in Greece proper: Pindos, Andania, Kelenderis, and Karandai in Aitolia
(SIG. 546. 14).
122 Thomsen RPLE, Marr JK, cf. Bleichsteiner 72, Hall CRPS, CGBA
292-3.
128 Kretschmer ASK, Sigwart 148-59.
V MATRIARCHAL PEOPLES OF THE AEGEAN 177
of Indo-European as a definite category may have to be revised.
Problems so far-reaching are not to be solved, or even ade-
quately stated, in a few pages. We must hold ourselves in
patience pending further progress in Anatolian prehistory.
Meanwhile I would merely insist that the ancient Greek
traditions concerning these early ^Egean peoples cannot be
dismissed as effusions of popular ignorance or antiquarian
speculation. When the pieces are put together they make a
coherent picture, which harmonises with the pattern emerging
from archeological and linguistic research.
5. The Minoans
The earliest known occupants of the Cyclades were settlers
from the east and south, perhaps largely from Crete, who were
acquainted with copper. This culture, known as Early Cycladic,
developed under Minoan influence. Early in the third millenn-
ium it spread to the Peloponnese, Central Greece, and southern
Thessaly (Early Helladic). The people that introduced it may
be identified with some confidence as the Carians and Leleges. 120
The neolithic population of Crete included an element from
North Africa. Their waistcloth and codpiece, and their
figure-of-eight shield, have parallels in Libya and pre-dynastic
Egypt. i3° But place-names of the type mentioned above are
commoner in Crete dian anywhere else outside Anatolia, and
the cult of die double axe survived in Caria after it had passed
into legend at Knossos. 131 For these and other reasons it is
agreed that the Minoan Cretans had affinities with the Carians,
Leleges, and Lycians.
These des have left their mark on the Greek tradition.
Sarpcdon, whom we meet in the Iliad as a grandson of Bcllero-
phon (p. 164), appears elsewhere as a brother of Minos, the
king of Knossos. 132 The first is the Greek version, the second
120 D. G. Hogarth in CAH 2. 555, Frodin 452. Almost ail the instances
of the place-name Minoa lie in the Caro-Lclcgian area: Amorgos (Nic.
Dam. 47, Androt. 19), Paros (Nican. 6), Delos (A. J. Evans PM 3. 74),
Laconia (Str. 368), Nisaia (Str. 391).
133 Hall CGBA 25-7.
131 For other Anatolian connections sec Pcndlcbury 42. "
132 Hdt. 1. 173.
M
V
178 STUDIES IN ANCIENT GREEK SOCIETY
Lycian and Minoan. A Cretan origin was assigned to the cult
of Zeus Atabyrios in Rhodes and the Carian settlement at
Miletos . 188 The Lycians, and the Carians of Kaunos, were said to
have come from Crete . 184
In these traditions the focal point is Crete. There is no hint
of a reverse movement to Crete from Anatolia. That is because,
thanks to their political supremacy, the Minoans dominated
the tradition. But their version was not undisputed. The
Carians insisted that their ancestors had reached the Aegean
islands from the Anatolian mainland, and in proof, they
appealed to their kinship with the Lydians, who had no
connection with the islands . 186
That Minoan civilisation was in some sense matriarchal is
generally acknowledged. One of the few facts about it, apart
from legends, that the Greeks remembered was that ‘in Crete
it had been customary for women to appear in public *. 136 The
custom impressed them because it contrasted with their
own. Not only did these women appear in public, but on the
frescoes, gems, and seals excavated by Evans we see them
strenuously engaged as boxers, bull-leapers, acrobats, chario-
teers, and hunters. 1 ” They even made pots . 138 In Greece we
never hear of a female potter in real life, and even in religion
only faint vestiges survive, such as the worship of Athena, as
patroness of the craft, and those curious girls of gold em-
ployed by Hephaistos in his smithy . 139 And yet the com-
parative evidence leaves no doubt that the art of baking clay
was invented by women . 110 These Minoan potilres supply the
link between Greek civilisation and primitive practice.
The Minoan rules of inheritance will not be known until
the inscriptions have been interpreted, but they are not likely
to have differed fundamentally from those we have found in
Lycia and other parts of the Near East . 141 The religious evid-
ence, which is relatively full, will be reviewed in Chapter VII.
133 Apld. 3. 2. 1, Eph. 32. 13 * Hdt. x. 172-3. 186 Hdt. 1. 171. 5-6.
130 Pto- Thcs. 19. Glotz CE 143. 138 A. J. Evans PM 1. 124-5.
139 Horn. Epig. 14, II. 18. 417-21.
wo Briffault I. 466-77, Mason 91-113.
1 41 At Gortyna in Greek times the son of a freewoman by a slave was
free if born in the mother's house: L Gort. 7. x, cf. p. 99.
V MATRIARCHAL PEOPLES OF THE AsGEAN I79
6 . The Hittites
We have completed our circuit of the Aegean, and on all
sides we have found vestiges of the matriarchate. But there is
still one people that demands attention.
The Hittites are believed to have entered Anatolia from th$
Caucasus. 142 They were a mixed stock, pastoral and warlike. 143
The use of iron was known to them at least as far back as the
thirteenth century. 144 One of their languages was Indo-
European. Their capital was Hattusas, the modern Boghaz-
keui, in N.W. Cappadocia. 146 They built up an extensive
empire controlling the whole of Cappadocia, a good part of
Syria, and some districts in central Anatolia. Further west,
Hittite monuments have been found at Sardeis, the Lydian
capital, on the heights of Sipylos, and down the Hermos valley
to the sea. 140 It has been suggested that the Atyadai, the first
Lydian dynasty, were subject to Hittite overlords. 147 Myrsilos,
the last of the Herakleidai, has the same name as Mursil, who
became king of the Hittites about 1350 b.c . 148 The third
Lydian dynasty, the Mermnadai, came from the country of the
Leukosyroi or ‘White Syrians’, who may have been Hittites. 140
Further, Tarchon or Tarquinius, ancestor of the Etruscans,
seems to be named after the Hittite wargod, Tarkhun. 160
Some of these equations are conjectural, but on the main
point there is agreement. At the height of their power the
influence of the Hittites extended down the waterways of the
Hermos and Maiandros to the ASgean.
The early Hittite kings were patriarchal and polygamous,
142 Cavaignac 14-5.
143 lb. 5, 42.
144 Cavaignac 4, Hall CGBA 253, cf. Str. 549, A. Pr. 740-1.
146 Cavaignac 1-2.
148 D. G. Hogarth in CAH 2. 264, 548, cf. Garstang 18, Lethaby 13.
147 Garstang 18.
148 Hdt. 1. 7; Hogarth in CAH 2. 264
140 Nic Dam. 49, Apld. 2. 5. 9; Garstang 171.
iso Kretschmer SLS 28. 104, 1 12-4, Blumel HT: Lydan *tar\u ‘be strong',
Gk. TapxfaJ, Hittite tarrl- be powerful’ (Sturtevant 153). Among the
proper names in Hittite documents are Tarkundaraba, Tarkulara, Tarkunazi,
Tarkumuva, cf. Str. 676 Tarkondimotos (Cilicia); Laroche 89.
V
l8o -STUDIES IN A'NCIENT GREEK SOCIETY
with succession from father to son, but in the days of their
greatness we find the queen and queen-mother in positions of
authority, the latter being associated with the king in his
official acts . 161 It would appear that their native institutions
were modified under Anatolian influence. So in religion. They
adopted Ishtar from Babylonia, Hepa and her consort Teshub
from the Mitanni . 168 Among the sculptural reliefs at Hattusas
is the figure of a female warrior, a goddess or priestess, perhaps
the queen her self in battle dress. In this warrior queen we may
recognise the prototype of the Amazons . 1 68
7. The Legend of the Amazons
The legend of the Amazons fascinated the Greeks. They
carried it with them wherever they went. It grew with their
own expansion until the whole of the known world had been
peopled with these romantic figures and their origin forgotten.
Their home, according to the prevalent tradition, was on the
north coast of Anatolia or further east in the Caucasus.
Herodotus relates how, after being defeated and taken prisoner
by the Greeks, they overpowered their captors and escaped -by
sea to the Crimea, where they became friendly with the
Scythians . 1 64 Later writers take them much further afield.
According to Diodoros they were natives of Libya. After
making themselves mistresses of that country they marched
under their queen Myrine to the western borders of the world,
fabulous Atlantis, where they overcame the Gorgons; then,
turning eastwards into Egypt, where they made an alliance
with Horus, the son of Isis, they fought their -way through
Arabia and Syria, subjugated the highlanders of Tauros, and
passed on through Anatolia to the Aegean coast, where they
founded several cities named after the bravest of their leaders.
Thence they made their way by Lesbos and Samothraike to
Thrace, and so, having conquered the world, they returned in
triumph to their Libyan home . 165
161 Cavaignac 52, 72, 85. zb. 1 16. wa zb. u6, Garstang 86-7.
164 Hdt. 4. no-3, cf. A.Pr. 7 A 9 SU Str. 505, 547, Pans. 1. 41. 7, Hp.
Aq. 17 .
185 D.S. 3. 52-4.
V MATRIARCHIAL PEOPLES OF THE AEGEAN iSl
Throughout the JEgcan area and along the north coast of
Anatolia there were local monuments called Amazoneia and
legends commemorating their adventures, but the region in
which they are said to have founded cities is more circum-
scribed. A number of these were on die shores of the Propontis
FIG. 7. Amafcn: Attic vac
and Paphlagonia . 166 The remainder were all on that part
of the /Egcan coast which was known later as Afolis and
Ionia — Myrine, Mytilcnc, Elaia, Anaia, Gryneia, Kymc,
Pitanc, Smyrna, Latorcia near Ephcsos, and Ephcsos itself,
which is said to have been ruled by an Amazon named
Smyrna. 1 ”
lscThiba (Am fr. 58), Sinope (Hec. fr. 352), Nifcaia (Oust, d D.P.
S2S, cf. Plu. Tics. 26), Amastris (Dcm. Bith. 9 *--FIIG. 4. 585), K)T.r.a
and MjTlcia=rApamcia (St. B. x.w.)
1 5 " D.S. 3. 54, Am fr. 5 S, Str. 550, 63 5, Sen-. d Verg. A. 4. 345, Ath. 5:0.
V
1 82 STUDIES IN ANCIENT GREEK SOCIETY
The ancient shrine of Artemis at Ephesos — Diana of the
Ephesians — was founded by Amazons . 158 This tradition has
been confirmed by excavations, which have brought to light
statutory groups of female hunters or warriors, evidently
votaries of the goddess like the Korai on the Athenian acropolis.
These monuments were published by Lethaby, who observed
among the early remains clear signs of Hittite influence.
Garstang agrees with him in associating the Amazons with a
Hittite cult, from which the worship of the later Artemis was
descended . 159
The current explanation of their name was that being
hampered in battle by their breasts they adopted the practice
of cauterising one or both in infancy, and so became known as
‘breastless* ( amazoi ). 160 Another view was that some women of
Ephesos, abandoning the natural vocations of their sex, took to
warfare and agriculture, and since they used to reap ( amdo )
with girdles (gonai) round their waists, they were called
Amazons . 181 We need not set much store by these etymologies,
but the idea behind the second is suggestive. It is from the
same point of view that they were identified with some
Caucasian tribes, in which, as reported by Strabo, 'the women
did all the ploughing, planting, pasturing, andhorse-breeding ’. 188
The same idea appears again in what Diodoros says of their
social life:
The Amazons were a people ruled by women, and their way of life was
very different from ours. The women were trained for war, being obliged to
serve under arms for a prescribed period, during which they remained
virgins. After being discharged from military service they resorted to men
for the sake of having children, but retained in their own hands the control
of all public affairs, while the men led a domesticated life just like the
married women in our own society. 188
To complete the picture we have only to add, on the
authority of Arrian, that they 'counted descent in the female
line 1 . 184
This myth was engendered, in its Greek form, as a symbol
for the matriarchal institutions of a theocratic Hittite settlement
168 Paus. 7. 2. 7, Tac. Ann. 3. 61. ieo Lethaby 10. 100 D.S. 3. 52.
lei Themistag. 3= FHG. 4. 512. 162 Str. 503-4. iea D.S. 3. 52.
UJ 4 Arr. fr. 58; Markwart 29.
V
MATRIARCHAL PEOPLES OF THE AEGEAN 183
at'Ephesos, dedicated to the Anatolian mother-goddess. 168
From there it spread over the AEgcan. Throughout the period
of Greek colonisation, which extended to all corners of the
Mediterranean, the legend continued to expand in response to
the expanding acquaintance of the Greeks themselves widi the
still matriarchal peoples with which they were everywhere
brought in contact, Or to put it another way, beginning as
handmaids of the warrior-queen at Hattusas, the Amazons
absorbed successively into a unified mythical concept all the
other matriarchal figures diat arose on the widening Greek
horizon — the Lydian Omphale, the Lemnian Hypsipylc, the
Assyrian Semiramis, die queens and quecn-modicrs of Egypt
and Ethiopia, Tomyris of the Massagctai, and die capable,
high-spirited women of coundess other primitive tribes in
Arabia, Libya, Italy, Gaul, and Spain. 166 The Amazons and the
women of Lemnos are polarised expressions of the same idea.
In the Lemnian legend the concept of modier-right has been
reduced to the level of a revolt against the later social order,
which, once established, claimed to be primeval; in die
Amazons it has been severed from reality, romanticised, freed to
float on a harmless flight of fancy.
8. The Mitiyai
We have now to consider what place in diis matriarchal
world can be assigned to die first carriers of Greek speech.
The infiltration of the new language must have begun far
back in die second millennium. If, as many believe, the im-
migrants came from the .Danube basin, they must have moved
168 The Hitticc mother-goddess was related to the Armenian, who in-
spired the legendary Semiramis. Ic is possible that in tracing the Amazons
to the Caucasus die Greeks were following a tradition which recognised
the Caucasian origin of Artemis. The place-name KizkaFah, Maiden's
Casde, is still a common one for hills surmounted with earthworks in
Armenia and Azerbaijan: C. F. Lchmann-Haupt in Roscher LGRM 4. 701.
166 The Nayars of Kerala preserved their matriarchate until after the
war of 1914.-18, and dieir women, 'whose beauty, self-respect, and elegance
arc proverbial, represent also a far healthier type than die Brahmin girls, i.e.
the patriachaliy ruled women of the same country . . . and have developed a
standard of intellect, character, and physical fitness equal to that of the
men’: Ehrcnfels 58-9-
V
184 STUDIES IN ANCIENT GREEK SOCIETY
down the Axios (Vardar) valley or else along the Adriatic
coast into Epeiros. In either case they would have been at-
tracted to the rich Thessalian plain watered by the Peneios and
its tributaries. Indeed, it has been proposed to identify them
with the neolithic culture named after the Thessalian site of
THE DIMINI CULTURE (Thessalianl) Map IV
Dimini. 167 These Dimini people were immigrants from the
north who established themselves in eastern Thessaly, with
extensions as far south as Corinth, where their remains have
been found overlaid by those of the Cycladic culture (Early
Helladic) mentioned above (p. 177). They fortified their
187 Hall CGBA 248. On the Dimini culture see Hansen 22-76.
V MATRIARCHAL PEOPLES OF THE AEGEAN 1S5
villages and introduced a new type of dwelling-house, the
‘megaron’. This identification is of course conjectural, but at
least it points to southern Thessaly and Bocotia, where the
earliest movements from north and south overlap, as a pro-
mising field for exploration.
Before making use of the traditional Greek genealogies it is
necessary to define so far as possible their historical value.
In a sense, as we shall sec, they are all fictions, but they cannot
be dismissed for that reason, because fictions arc significant.
The sons of Hellen — Aiolos, Doros, and Xouthos the father of
Ion — are palpably fictitious in the sense that no such persons
ever existed. They embody the national self-consciousness of
the Greeks — their sense of unity as Hellenes and of diversity
as AEolic, Doric, Ionic — and that is not fiction but fact.
In primitive society the elders of die clan carry in their heads
a fully articulated pedigree covering all the living members
and as many of the deceased, together with their marriage
connections, as arc needed for transmitting the clan’s tradi-
tions and regulating its conduct. But as time passes the clans-
men of the past lose their individuality, merge into one
another, and fade into the generalised concept of clan ances-
tor, who stands to other figures of the same kind as brother
or cousin according to the manner in which the clans have
evolved. The chronology tends to be foreshortened, but the
sense of origin remains.
Such traditions retain their vitality as long as kinship re-
mains the dominant factor in social life. When the tribal
system breaks up, they become stereotyped, and as the class-
struggle develops they become exposed to arbitrary reconstruc-
tions and distortions. It is these later redactions that contain
the main sources of error. Where the genealogies have re-
mained relatively undisturbed, as among the Icelandic Norse-
men and the Maoris, they are. within limits, remarkably
accurate. The Greek pedigrees, however, belong to a more
advanced stage, and the margin of error is accordingly wider.
On the other hand, the very diversity of the Greek tradition,
resulting from the autonomy of the city-states, provides material
for analysis like the variant readings of different manuscripts.
Chadu-ifk GL 1. 270-6, 5. 222-5.
V
l86 STUDIES IN ANCIENT GREEK SOCIETY
The historical value of a tradition is not necessarily annulled
by the fact that it has come down to us in different versions.
Of two variants both may be valid, even though they are con-
tradictory. Boutes was a son of Poseidon; Boutes was a son of
Pandion (p. 126). Neither of these statements is a fact. The
one symbolises the ultimate origin of the Boutadai, which' will
be investigated in a later chapter; 169 the other their admission
to the cults of the Erechtheidai. We may be sure that, when
they took over the worship of Athena Polias and Erechtheus,
they did in fact submit to some ceremony of affiliation or
adoption; and according to tribal ideas the introduction of a
strange clan involves an adjustment of the pedigrees as a
formal register of the act of rebirth by which the union has
been effected.
One feature of the Greek pedigrees strikes us at the first
glance. From the point where they emerge into the full light
of history women are mentioned quite frequently. This is
largely because, being recent, the details are fully remembered.
Besides, even under the democracy, the old families retained
a good deal of their prestige, and sometimes their intermar-
riages had a political significance. But in the preceding period,
as far back as the Dorian conquest, women’s names are con-
spicuous by their absence. The main purpose of the genealogies
appertaining to this period was to preserve the line of clan
descent for the sake of its accompanying privileges, and, since
descent was patrilineal, the women were a negligible factor.
But then, going still further back, we find women more pro-
minent than ever. Take the stemma of the Kodridai, to which
Solon and Plato belonged. 170 It covers thirty-two generations,
from the fourteenth century to the fourth. In the first three
the wife’s name is recorded in almost every case, and in several
cases in the fourth; but after the fourth generation there are no
more women till we reach the thirtieth. Some of these early
women’s names are mere names, with no apparent functional
value; but they must once have been more than that, or they
would not have impressed themselves so deeply on the tradi-
tion. Our greatest difficulty in interpreting these prehistoric
pedigrees is that they have been transmitted to us through a
lea See pp. 265-6. 170 The stemma is given by Petersen 94.
V MATRIARCHAL PEOPLES OF THE JEGEAN 187
period in which the woman’s part in deter mining succession
and descent had ceased to be understood.
The city of Orchomenos, called the Minyan Orchomenos to
distinguish it from others of the same name, lay a little to the
north of the point where the Kephisos empties into Lake
Kopais. 171 It is the most northerly site on the mainland at which
Minoan culture was securely established. From the earliest
times Orchomenos had disputed with Thebes, another
Minoan centre, the control of the Boeotian plain. Their
rivalry lasted down to 364 B.C., when Orchomenos was sacked
and its people sold into slavery. Its traditions, save for a few
fragments, perished with it. The Thebans had triumphed.
Even so they were unable to efface the memory of a time when
their own city had been ruled, perhaps even founded, by kings
of Orchomenos. 172
The first king of Orchomenos was Andreus, a son of Peneios.
During his reign a newcomer, Athamas, was allotted lands on
Mount Laphystion and on the lakeside at Koroneia and
Haliartos. Andreus married a granddaughter of Athamas, and
had a son, Eteokles, who succeeded him. 173 In his reign Almos,
son of Sisyphos, entered the country and settled at a village
which was named Almones after him. Almos was succeeded by
his daughter’s son, Phlegyas, and he by Chryses, son of his
mother’s sister. The Phlegyai were a warlike people and
ravaged the country as far as Delphi. They were destroyed by
thunderbolts and earthquakes.
Then a new dynasty began, founded by Minyas, son of
Poseidon, a ruler of fabulous wealth, which he stored in sub-
terranean treasuries. 174 His son was Orchomenos. The next king
was Klymenos, a great-grandson of Athamas. It was his son,
Erginos, who conquered Thebes. Trophonios and Agamedes,
171 This account of the dynasties of Orchomenos is from Paus. 9. 34-7;
the principal variants are given in the footnotes.
172 Apld. 2. 4. 11, D.S. 4. 10. 3-5, Paus. 9. 37; Oi. 11. 263-5.
173 Eteokles is also given as father of Minyas and Orchomenos: Pi. I . I.
79 sch.
174 Minyas is variously described as a son of Poseidon by a daughter
of Aiolos (Pi. P. 4. 120 sch.) or Okeanos (Pi. O. 14. 5 sch.) or Boiotos
(A.R. 1- 230 sch.) or Hyperphas ( Od . 11. 326 sch.), or as a son of Orcho-
menos, Eteokles, Aleos, or Ares (Pi. L I. 79 sch.).
V
l88 STUDIES IN ANCIENT GREEK SOCIETY
the sons of Erginos, were famous architects of shrines and
treasuries. The kingdom then passed to Askalaphos and
lalmenos, whom a great-granddaughter of Klymenos had born
to Ares. These led the contingent from Orchomenos to the
Trojan War.
Table IX
THE KINGS OF ORCHOMENOS
Peneios
I
Andreus = Euippe
Eteokles Almos
Chryse Chrysogeneia
Minyas Phlegyas Chryses
Orchomenos Klymenos
Erginos Azeus
Trophonios Agamedes Aktor
Astyoche
Askalaphos lalmenos
These pedigrees are confused, incoherent, and conflicting.
They represent the attempts of antiquaries to square a local
tradition surviving only in fragments with the Homeric poems
and other literary sources. Even so it is possible to disentangle
the guiding thread.
Peneios, father of the first king, is the river that flows
through the Thessalian plain. Almos, the eponym of Almones,
also left his name in a Thessalian village variously known as
Almos, Salmon, Halmonia, Salmonia . 176 It lay near the Thes-
salian Orchomenos (the later Krannon), which at one time,
175 plin. NH. 4. 29, St. B. MivOa, Hell. 27.
V
MATRIARCHAL PEOPLES OF THE AEGEAN 189
wc arc told, had been called Minycios . 170 Sisyphos, die father
of Almos, was located at Corinth, but Aiolos, the fadicr of
Sisyphos, was a native of Thessaly . 177
So with the variants. Phlcgyas appears elsewhere as a son of
Antion, but Antion was a grandson of Lapithcs, eponym of a
Table X
Pcneios
THE LAPITH AI
r~
Atrax
Elmos
!
Kaineus
1 I
Hvpscus Stilbc-- Apollo
Lapithcs
r
Koronos Astyagy ia <- Pcriphas
Andntimon Amion
Phorbas
Augens
Thoas Phlcgyas Ixion Phylcus Agasthcncs
, I ll
Haimon Pcirithoos Mcgcs
Polyxcnos
Oxylos
Polypoitcs
Akror
Eurytos
Thalpios
Thessalian tribe. 17 " One of his brothers was Gyrton, a town in
N.E. Thessaly above the Vale of Tempe . 170
Another son of Pcneios was Atrax, a town further up the
valley . 180 Kaincus, grandson of Atrax, was a famous Lapith
chief . 1 81 His father was Elatos, eponym of Elatcia, which lies in
the same valley below Gyrton . 1 82
170 Plin. Nil. 4. 29.
177 Apld. 1. 9. 3, 1. 7. 3* Sisyphos himself is described as a native of
Thessaly: see below n. 223*
i?8 Phlcgyas was a brother of Ixion (Str. 442), the son of Antion (A.
fr. 89), the son of Pcriphas, die son of Lapithcs (D.S. 4. 69). There are
other variants but the Lapith connection is constant.
170 St. B. roprwv. 180 St. B. "ATpa?. wi Ant. Lib. 17. 182 Die. 30.
V
190 STUDIES IN ANCIENT GREEK SOCIETY
No matter how these traditions may contradict one another
in details, they concur in indicating that Orchomenos was
occupied in early times by branches of the Lapithai, which had
reached Bceotia from N.E. Thessaly; and what is more, their
Thessalian homeland, as defined by the eponyms, coincides
with the environs of the modern Rachmani, which are ex-
ceptionally rich in remains of the Dimini culture. 1 83 It is pos-
sible therefore that the neolithic culture of the Rachmani
district corresponds to the Lapithai of Greek tradition.
Aiolos appears in the Homeric poems as the father of Sisy-
phos and Kretheus. 184 In Hesiod the same paternity is claimed
for Athamas, Salmoneus, and Perieres. 186 Later writers extend
it still further. Like Doros and Ion, who are unknown to
Homer, Aiolos is a relatively late concept. He symbolises one
of the three branches into which the Greeks found themselves
divided when they settled down in their new home. For this
reason he cannot be relied on for the early history of the tribes
and clans affiliated to him. In spite of this, the fact that he was
assigned to a Thessalian origin is significant, and at least two
of his sons, Sisyphos and Kretheus, have independent ties
with the same region. As we have seen, Almos, son of Sisyphos,
bears a Thessalian name. Sisyphos himself reigned at Ephyra,
identified as Corinth, but there was another Ephyra in Elis
and a third in Thessaly. * 88 This should mean that emigrants
from Thessaly had settled in Corinth and Elis, and we shall
find that such was in fact the case. Sisyphos, it will be recalled,
was the grandfather of Bellerophon, from whom the Ionian
kings were descended — an indication, as I have pointed out,
that his stock was Greek-speaking (p. 165). Kretheus was
the founder of Iolkos at the head of the Gulf of Pagasai. 187
His wife was Tyro, who bore him three sons — Aison, Pheres,
183 Hansen 26-8, 33-7, 43-4, 50-5, 78-113, 182-4.
184 ll. 6. 152-4, Od. 11. 235-7.
186 Hes. fr. 7.
188 IL 6. 152-3, Apld. 1. 9. 3, Str. 328, 333, 338. The Thessalian
Ephyra was the later Krannon (Str. 338, 442). There was a fourth near
Dodona, founded from Thessaly (Pi. N. 7. 37, Str. 324) and a fifth in Aitolia
(Str. 338) where there was a Lapith colony, represented in Homer by Thoas
(ll. 2. 638; see Table X).
187 Apld. 1. 9. xi.
V MATRIARCHAL PEOPLES OF THE jEGEAN 191
and Amythaon. 188 Aison stands for Aisonis, another setde-
ment on the Gulf; Pheres founded Pherai in the same district. 189
Aison's son, Jason, the leader of the Argonauts, set sail from
Iolkos. 1 " I have already alluded to his sojourn in Lemnos,
where he begot Euneos of the Euneidai (p. 175). Amythaon,
the third son of Kretheus and Tyro, was the father of Mel-
ampous and Bias. These migrated to the Peloponnese, where
Melampous married a daughter of Proitos, the king of Argos
who entertained Bellerophon. 191 In Elis there was a stream
called the Minyeios, in which Melampous purified the
daughters of Proitos after Dionysus had driven them mad. 192
From him was sprung the priesdy clan of the Klytidai, who
administered the Olympic Games. 193
So far we have not established any direct connection between
Sisyphos and Kretheus beyond their affiliation to Aiolos, which
we have decided to disregard. But there is still one small
detail. Sisyphos is said to have had children by Tyro, who
killed them at birth. 1 ® 4 This looks like a reminiscence of an
ancient tie between Sisyphos and Tyro, which the Corinthian
tradition suppressed.
Tyro was also at home in Elis, where she appears as a daughter
of Salmoneus, eponym of Salmone to the north of Olympia. 196
There she became enamoured of the River Enipeus, and, either
to him or to Poseidon disguised as the river, she bore twin sons,
Pelias and Neleus. 1 ** Pelias ‘dwelt in Thessaly’, where he
begot Alkestis; N eleus went south to the Messenian Pylos,
where we meet his son Nestor in the Odyssey , 197 When the
188 Od. 11. 235-9. 189 P^er. 5 ®* ApM. I. 9. 14. 190 Apld. i. 9. 16.
191 Apld. 1. 9. II, 2. 2. 2. 192 II II. 722, Paus. 5. 5. 7, 5. 6. 3.
193 Paus. 6. 17. 6. Other descendants of Melampous and Bias can be
traced in Megara, Messenia, and Akarnania (Paus. 1. 43 * 5 , 4 * 34 * 4 *
Hdt. 7. 221). At Mantineia there was a shrine of Poseidon Hippios erected
by Trophonios and Agamedes (Paus. 8. 10. 2), also tombs of the daughters
of Pelias (Paus. 8. 1 1. 2). These cults may be referred to the Tyroidai or
Lapithai.
' i94 Hyg. F. 60, 239- Neleus was said to have been buried secretly at
Corinth: Paus 2. 2. 2.
1 96 Apld. 1. 9. 7-8, Str. 356. i®8 Od. n. 235-59.
197 od. 11. 281-6, Apld. 1. 9. 9, Paus. 4. 2. 5. That the residence of
Nestor was the Messenian Pylos, not the Triphylian, has been confirmed
by recent excavations: Blegen EP.
V
192 STUDIES IN ANCIENT GREEK SOCIETY
Dorians broke into the Peloponnese, these Neleidai of Pylos
fled to Attica, where they founded some of the most renowned
of die Athenian clans— die Alkmeonidai, Peisistratidai,
Paionidai, and Kodridai .* 98 The Kodridai led the migration to
Ionia, where they were installed in several cities as kings, like
their distant relatives, the descendants of Bellerophori from
Lycia . 189
This tradition is important from several points of view. In
the first place, it is hard not to believe that a series of migra-
tions so extensive in space and time was part of the movement
that established the Greek language in its historical domain.
There is a strong presumption that the stock of Kretheus, as of
Sisyphos, spoke Greek.
Secondly, the focus of the whole movement is Thessaly.
Salmone, the abode of Salmoneus in Elis, is only another form
of the Thessalian Almos . 200 The River Enipeus, which Tyro
loved, appears in Thessaly as a tributary of the Peneios . 201
The stream Minyeios harks back through the Minyan Or-
chomenos to the Thessalian Orchomenos or Minyeios. Con-
trariwise, there was a river in Elis called the Peneios . 202 And the
story of Bias, who, before he could wed the lovely daughter of
Neleus, was sent to fetch the catde of Phylake, had evidently
been transferred from Thessaly, because Phylake lies between
the Gulf of Pagasai and the Thessalian Enipeus . 203
Then there is Tyro herself. The daughter of Salmoneus in
Elis, the wife of Kretheus in Thessaly, hers is the name that
unites the two branches of the stock. She is the common an-
cestress, the first mother of the clan. Can it be that Salmoneus
and Kretheus have been inserted at the head of the tree in
order to adapt to the ideas of a later age a tradition of matri-
lineal descent? With this possibility in mind, let us turn to the
Minyai.
138 Hdt. 5 ’ 65. 4 > Pa us. 2. 18. 8. Some descendants of the Neleidai
survived in Messenia: Str. 355.
139 Hdt. X. 147, 9. 97. Herodotus describes the Kodridai as Kaukones,
which I take to mean that their followers included Kaukones from Pylos.
200 See above n. 175.
201 Str. 356, 432, cf. Apld. 1. 9. 8.
202 Str. 337-8.
2 03 Od. xx. 287-97, Apld. 1. 9. i2, Str. 433, 435.
V
MATRIARCHIAL PEOPLES OF THE AEGEAN I93
The Minyai were the people of the Minyan Orchomenos, so
called after Minyas, who refounded the city. He seems to have
come from the Thessalian Orchomenos, the former Minyeios,
but, though hailing from Thessaly, he has no ties with the
Lapithai or Tyroidai. He marks the intrusion of a new element.
We turn to archeology for the clue.
Shortly after 2000 B.C. Orchomenos was destroyed and re-
occupied by a people using a distinctive type of pottery known
as 'Minyan ware*. This is the name Schliemann gave it, and
perhaps it was truer than he knew. Pottery of this type has
been found in Thessaly, Macedonia, and Troy, with secondary
extensions into central and southern Greece, where it overlays
Early Helladic (p. 177). It is now many years since Forsdyke
argued that it reached Greece from Troy, and Heurtley’s
recent excavations in Macedonia support this view. 204 Heurtley
considered that it was developed in Thessaly and Central
Greece by immigrants who had come by way of Macedonia
from N.W. Anatolia. Further, he postulated a common pro-
venance for this and the Early Helladic culture 'somewhere
east of Troy’. I would suggest, with all due reserve, that, just
as Early Helladic was the work of the Carians and Leleges
(p. 177), so Heurtley’s immigrants from N.W. Anatolia are
die Pelasgoi.
Under the Minyai Orchomenos was drawn into the orbit of
Minoan Crete. This epoch, culminating in the great palace
excavated there, which dates from about 1400 B.c., is the one
to which we may refer the architectural feats of Trophonios
and Agamedes and the traditions concerning the daughters of
Minyas. These are of special interest because they testify to
intimate relations with the Tyroidai.
Minyas had a bevy of daughters. Their names were Kly-
mene, Periklymene, Eteoklymene, and Phersephone. 206 Since
Klymenos was a title of Hades, 206 and Persephone his queen,
they point to a Minoan palace cult of Demeter-Persephone,
like the one founded by Kadmos at Thebes (p. 124).
204 Forsdyke PMW, Heurtley PM 118-23.
206 A. R. 1, 230 sch., Pher. 56.
®o6 Las. ap . Ath. 642e=Diehl 2. 60, Call. fr. 139 Mair, cf. Paus. 2.
35.4, and see H. W. Smyth 300.
N
194 STUDIES ltf AtfCIEtfT GREEK SOCIETY
Table XI
minvas and tyro
MINYAS Kretheus= TYRO =Enipeus
r 1 i ~i 1 I
Phersephone=Iasos Klymene=Phylakos Periktymene — Pheres
1
Niobe= Amphion
Eidomene=Amythaon
r- 1 -.
Meiampous Bias
1 .
Alkimede = Aison
I
Hypsipyl e =Iason
Euneos
1 .. r
Chloris= Neleus
I
Nestor .
r
Phyiomache=PeIias
I 1
Admetos=Alkestis
I
Eumelos
Phersephone was die mother of Amphion , 207 who had two
daughters, Chloris and Phylomache. These married Neleus and
Pelias, Tyro’s sons by Enipeus . 8 08 Periklymene married Pheres,
207 pher. 56. Amphion is also given as a son of Antiope, granddaughter
of Hyrieus, i.e. Hyria under Mount Kithairon: Apld. 3. 5. 5.
sob oi. xi. 281-2, Apld. 1. 9. 10. Neleus is said to have ruled Orchomenos
as well as Pylos (Pher. 56), presumably by right of marriage to Chloris.
The wife of Pelias is also given as Anaxibia, daughter of Bias (Apld. 1. 9. 10).
V
MATRIARCHAL PEOPLES OF THE AEGEAN I95
one of Tyro’s sons by Kretheus. They had two children —
Admetos and Eidomene . 200 Admetos married Alkestis, daughter
of Pelias and Phylomache. Eidomene married Amythaon,
Tyro’s second son by Kretheus, to whom she bore Melampous
and Bias . 210 Klymene married Phylakos, and their daughter,
Alkimede, married Aison, Tyro's third son by Kretheus, to
whom she bore Jason . 211 All this sounds remarkably like a
tradition of two intermarrying clans.
The Minyai were properly the people of Orchomenos. But
Jason and the Argonauts are also described as Minyai. Why
should the inhabitants of Iolkos have been designated by this
name? One explanation offered in antiquity was that Minyai
from Orchomenos had settled at Iolkos . 212 That is likely
enough, because S.E. Thessaly has yielded fairly plentiful
Mycenean (Late Helladic) remains, poorer than the Boeotian
and introduced from that direction . 313 Another, given by
Apollonios, the learned author of the Argonautika, was that the
Minyai of Iolkos were so called because their leaders were
sprung from the daughters of Minyas . 214 In other words these
descendants of Tyro, who had settled round the Gulf of
Pagasai and intermarried with the dynasty of Orchomenos,
were Minyai in the female line.
The same conclusion is reached by approaching the problem
from an entirely different angle.
At the Boeotian festival of the Agriania a band of women was
pursued with a drawn sword by die priest of Dionysus, who
was entitled to kill the hindmost if he caught her. The ex-
planatory myth referred to the daughters of Minyas. After
refusing to be initiated into the mysteries of Dionysus they
were seized with a mad desire for human, flesh. They cast lots,
and the sister on whom the lot fell gave her own child to be
torn in pieces and eaten. After that they ran wild in the
mountains, feeding on ivy, yew, and laurel . 216 The last detail
corresponds to another feature in the festival at Orchomenos.
209 A.R. 1. 230-3 et sch., Apld. 1. 9. 14. 210 Apld. 1. 9. xo-x.
211 A. R. 1. 45-7, 230-3. There were several other variants of Jason’s
paternity: Roscher LGRM 1. 197.
212 Str. 414. 213 Hansen 107. 214 A.R. 1. 229-32.
216 Pin. M. 299c, Ant. Lib. 10.
V
196 STUDIES IN ANCIENT GREEK .SO CIETY
‘The women’, says Plutarch, ‘fall upon the ivy in their frenzy,
tear it in pieces, and devour jt .’ 216
This ritual belongs to a well-known type. A human victim is
driven out into the country and there sacrificed as a scapegoat
for the sins of the community . 217 The pompai or processions of
Greek religion, in which the god’s image was escorted out of
the town and brought back again after a sacrifice, were rites
of the same nature . 218
The Agriania was also observed at Argos, where it was
associated with the daughters of Proitos. When Dionysus came
to Argos, the women refused to be initiated. The god drove
them mad, whereupon they killed the babes at their breasts
and devoured them. The daughters of Proitos, in particular,
roamed in distraction all over the Peloponnese, pursued by
Melampous at die head of a band of young men performing
an ecstatic dance. During the pursuit one of them died. We
are not told that Melampous killed her, but it sounds like it.
Eventually the survivors reached the River Minyeios near
Olympia, and there they were purified by Melampous, who
took one of them to wife . 219
It is clear from the identity of the two myths that the Argive
Agriania was based on the same ritual as the Boeotian . 220 We
may conclude that it was introduced into the Peloponnese by a
branch of the Tyroidai, represented in the genealogies by
Melampous, who had inherited it from the Minyai of Or-
chomenos through the female line.
The Attic Euneidai, descended from Jason, also had a cult
of Dionysus, associated with Dionysos Kittos, the Ivy Dionysus.
It was characterised by flute-playing and dancing, and one of
the clan’s privileges was to supervise the state processions
(pompa /). 221 This too, it seems, goes back to Orchomenos,
210 PIu. M. 291a. 217 Frazer GB-S. 218 G. Thomson AA 166-7.
210 Apld. 2. 2. 2, Hdt. 9. 34, Str. 346, D.S. 4. 68, Paus. 2. 18. 4, 5.
5. xo, 8. 18. 7.
220 The month Agrianios (=Att. Thargelion, SIG. 1031 n. x) occurs in
Bccotia, Sparta, Rhodes, Kos, Kalymnos, and Byzantium: Paton 327-30,
Fameli CGS 5. 300. This distribution agrees with the hypothesis that it
•was introduced to the Peloponnese from Boeotia: see G. Thomson GC 56.
82 i Paus. x. 31. 6; see p. 122 n. 91,
V ' MATRIARCHAL PEOPLES OF THE AsGEAN I97
where the ancestors of the Euneidai had acquired it as Minyai
in the female line.
Were these Minyai Greeks? Nilsson believes they were —
Ionian Greeks . 222 But I think we must draw a distinction.
Though the Minyai intermarry with the Tyroidai, there is no
ancestral connection. Accordingly, I prefer to regard the
Minyai of Orchomenos as non-Hellenic — perhaps, as I have
suggested, Pelasgian. But the Minyai of Iolkos are different.
If the stock of Tyro was Greek, then so were these — Minoan-
ised Greeks. But I should hesitate to call them lonians, because
it is unsafe to assume that Ionic existed at this early period as a
separate dialect. More probably their speech was die parent
of the later Ionic and AEolic. That would explain why, while
their descendants in Attica and Ionia became affiliated to Ion,
they were themselves always reckoned as descendants of Aiolos.
It is tempting to go further. I have suggested that the
Lapithai of northern Thessaly — Atrax, Gyrton, and Elatos —
are to be connected with the Dimini settlements around
Rachmani, which is just north of Elateia. Outside this dis-
trict, Dimini remains are most plentiful in the lowlands
round the Gulf of Pagasai. This was the homeland of the
Tyroidai. Dimini itself lies close to the ancient Iolkos. Further,
it appears that the Tyroidai had reached the Gulf from the
north; for the god whom Tyro loved, the father of Pelias and
Neleus, is described as Poseidon Petraios, referring to Petra
on the northern foothills of Olympos . 223 Were the Lapithai
and Tyroidai two branches of the Dimini people? Against this
it might be urged that in the genealogies the Thessalian
Lapithai are affiliated to Apollo, but that may have been due
to later influences from Delphi, and in a later chapter we shall
find their descendants, the Lapithai of Attica and the Pelopon-
nese, connected with Poseidon, and in particular with Poseidon
Petraios . 224 If the Dimini culture has left any trace at all in
222 Nilsson MOGM 155* HM 96.
223 pi. p. 4. 138. The shrine of Poseidon Petraios is said to have been
founded in memory of the birth of Sisyphos, whom the god produced in
the shape of a horse with a blow of his trident on the rock: EM. 'Irnnos 6
nooEiScbv.
224 See pp. 264-5.
V
198 STUDIES IN ANCIENT GREEK SOCIETY
Greek tradition, the Lapithai and Tyroidai have a strong claim
on it. And if we accept it, we may say that these were the
people who after assimilating the culture of Orchomenos
transmitted ‘Minyan ware' to southern Greece.
One point has still to be cleared up. Our conclusion is that
these Greek-speaking villages fringing the Gulf had been
drawn into intercourse with the urban culture of Orchomenos.
But this culture reached its zenith about 1400 B.c. Tyro, on
the other hand, if we accept the traditional chronology, can-
not be put further back than 1300. She is much too late.
Difficulties of this kind will confront us repeatedly in the
Greek genealogies. In a later chapter I shall argue that the
traditional chronology cannot be accepted as it stands. 226
Meanwhile it may be observed that our attitude to the chrono-
logical framework of these pedigrees must depend on our
estimate of their content. If we accept the traditional date of
Tyro at its face value, we accept her as a real person. I doubt
if anyone is prepared to do that. But, if she was not a real
person, she cannot have had a date. As soon as we recognise
her as merely a symbol for the common matrilineal origin of a
group of Greek clans, the chronological difficulty disappears.
"When the earliest remembered ancestors of the Kodridai and
Alkmeonidai emerge out of the mists of the Enipeus in chal-
colithic Thessaly, they fall under the spell of the matriarchal
priest-kings of Orchomenos, with whom they trade and marry.
But they are already a warlike people, established in a ring of
strongholds round the Gulf of Pagasai, and soon, taking to the
sea, they open up relations with Lemnos and Troy, and from
there, perhaps with Pelasgian pilots, they find their way to
Kolchis at the far end of the Black Sea; then, turning south,
they found kingdoms on the west coast of the Peloponnese. By
this time, it may be, they are becoming patriarchal again, but
later still, when their descendants cross the ABgean to Ionia,
they find themselves in a situation not unlike that of their
ancestors at Iolkos — compelled to marry Carian wives and to
construct their new patriarchal city-states in conscious oppo-
sition to the vast matriarchal world that stretches from their
doorsteps into the heart of Anatolia.
225 See p. 409.
V MATRIARCHAL PEOPLES OF THE AEGEAN I99
9. Some Matriarchal Survivals
It is to be feared that even those of my readers who suc-
ceeded in direading their way through the classificatory system
of relationship will be suffering from the effects of these
genealogies. This is regrettable, but it cannot be helped. If we
want to understand how we have become what we are, we
must train ourselves to see things from the standpoint of our
remote ancestors, who, since their lives were determined by
the supreme fact of kinship, saw kinship everywhere. We
must develop a genealogical oudook on the world.
By way of relaxation let us conclude the present chapter by
enquiring whether there can be found in any part of Hellas
‘some little town by river or sea-shore’ in which mother-right
lasted long enough to enter the light of history.
Lokroi Epizephyrioi was a Greek colony on the toe of Italy.
It was founded in 683 B.c. from Lokris in Central Greece,
whose early inhabitants are described by Aristode as Leleges. 228
Speaking of die colony as it was in the first century B.c.
Polybius says: ‘All dieir ancestral honours are traced through
women, as for example the noble rank enjoyed by descendants
of the Hundred Families.’ 22 * From other sources we learn that,
like the Lydians and Etruscans, the people of Lokroi practised
pre-nuptial promiscuity, 228 and that theirs was the first Greek
state to codify its laws. 220 Their retention of matrilineal succes-
sion is thus explained by the formal stabilisation of their
institutions at an abnormally early date.
The Hundred Families were descended from the original
nobility of the homeland, and we are told that what drove the
colonists to emigrate was the scandal caused by these high-
born ladies in consorting indiscriminately with slaves. 230 In
220 Arist. ft. 56o=Str. 322. Which Lokris he meant is uncertain (Str.
259).
222 Pib. 12. 5. 6. That &irt> tSv yovwK&v means 'through women', not
merely ‘from women', appears from Arr. fr. 58.
228 Clcarch. 6.
220 Arist Pol 1274a. 6-7, Eph. 47, Pi. O. 10. 17 sch. Under this code
alienation of the ancestral estates was illegal: Arist. Pol 1266b. 6.
230 D.P. 365-7, Plb. 12. 6b. The 'slaves' were doubtless serfs: Sotflios
covers both.
V
200 STUDIES IN ANCIENT GREEK SOCIETY
their defence it might be pleaded that they were no worse than
Omphale or* Tanaquil, or the Bantu queens of Africa; and
when the tradition goes on to stigmatise the colonists them-
selves as a ruffianly gang of runaway slaves and adulterers, we
smile again at the prejudice that created the crimes of Lemnos.
Not so, however, our learned historians. ‘That a body of
colonists formed of such unpromising materials*, we are
gravely assured by Grote, ‘should have fallen into much law-
lessness and disorder, is no way surprising, but these mischiefs
appear to have become so utterly intolerable in the early years
of the colony as to force upon everyone the necessity of some
remedy: hence arose a phenomenon new in the march of Greek
society — the first promulgation of written laws.’ 231 When we
reflect on what this theory of the origin of legislation implies —
that the most abandoned criminals are the natural leaders in the
onward march to law and order — it is hard not to laugh outright.
Lokroi Epizephyrioi may have been exceptional in making
such a bad start, but in this region there were other colonies
founded in the same period and in conditions economically
similar, if not morally.
Taras (Tarentum), on the heel of Italy, was even older than
Lokroi. It was founded by some men from Sparta called Par-
theniai, ‘maidens’ sons'. When the Spartans were conquering
Messenia, which took them many years, their wives consoled
themselves in the same manner as Clytemnestra, and these
‘maidens’ sons’ were the result. Their fathers are described as
Spartans who had stayed at home. At the end of the war the
husbands returned and punished the seducers by degrading
them to serfdom. 232 Such was the story, but it is not quite con-
vincing. If the culprits were freeborn Spartans, they were
doing what was not an offence in the fourth century (p. 143),
and so we may doubt if it was in the eighth. Moreover, the
reason why they were called ‘maidens’ sons’ can only be that
231 Grote 3. 378.
232 Theop. 190, Ant. 14, Serv. ad Verg. A. 3. 551. In another version
the fathers of the Partheniai are soldiers sent home on purpose to beget
offspring (Eph. 53); but they are described as tncwcxicrot, 'additional
bedfellows — a term implying a recognised class of serfs with special privileges,
like the K<xrcovaKo<p6poi of Sikyon (Theop. I95). In II 16. 180 irapStvios means
'son of an unmarried mother.'
' V MATRIARCHAL PEOPLES OF THE AEGEAN 201
their mothers were unmarried. The truth seems to be that
their fathers had been serfs all the time — serfs with a privileged
status, like those who enjoyed the favours of the Locrian
ladies. Hitherto the offspring of such unions had possessed a
claim on the maternal estate; but now, with the rich plain of
Messenia ready to be appropriated, the old rule was abolished,
and so those who were only Spartans on the mother’s side were
compelled, ‘bearing their birthright proudly on their backs’,
to seek new fortunes overseas. The foundation of Taras, as of
Lokroi, was an incident in the conflict that was going on in
the mother-country over rights of succession to real estate — the
struggle for the land. 233
Kypselos, the first tyrant of Corinth, seized power in 657 B.c.
He belonged to the clan Kaineidai, descended from Kaineus
the Lapith (p. 189). 234 He was born a few miles from Corinth
in the townland of Petra, which reproduces the Petra on the
foothills of Olympos. 236 Before his time the city had been ruled
by die Bakchidai, who claimed to have come in with the
Dorian conquerors — where from, we do not know. At first they
had ruled as kings, and, after the kingship was abolished, they
retained power through annual magistrates appointed exclu-
sively from themselves. 230 This clan observed the custom, as
Herodotus expresses it, ‘of marrying and giving in marriage
among themselves’. Here we have a confirmation of our view
that die early Greek clan was normally exogamous. Why then
were the Bakchidai endogamous? The historian continues:
One of them, Amphion, had a daughter, Labda, who was a cripple, and
since none of the Bakchidai would take her to wife she was married to
Eetion, son of Echckrates, of Petra, a Lapith of the Kaineidai. 23 *
Shortly afterwards the Bakchidai received from Delphi an
enigmatic oracle which they eventually construed to mean that
233 At the beginning of the war, asked why he wanted to fight his brother
Dorians of Messenia, the Spartan king replied, ‘I am going to enter
on our unallotted heritage’ (Plu. M. 23 1 e), i.e. divide the land. On another
occasion the Spartans were encouraged to invade Tegea with the promise of
‘a .fine plain to measure with the rope’ (Hdt, 1. 66).
234 Hdt. 5. 92P.
ess pj, p, 4. 246 sch.
230 Paus. 2.4. 3. There were over 200 of them (D.S. 7, Wesseling 4.15).
23 * Hdt. 5. 92P.
202 STUDIES IN ANCIENT GREEK SOCIETY - V
the son of Eetion would be the ruin of the city; and so, when in
due course Labda was delivered of a fine boy, they decided to
destroy it. Ten of them proceeded to the house of Eetion,
ostensibly on a friendly visit. Their plan of action, arranged
on the way, was that whoever got hold of the child first was to
dash its brains out. Labda suspected nothing and handed the
baby to one of them. At this moment it happened to smile.
This touched the murderer’s heart, and he passed it on to one
of his companions. Overcome with compassion in his turn, he
did the same, and so the beaming infant was bandied from one
to another and back into its doting mother’s arms. After
leaving the house the soft-hearted assassins broke into mutual
recriminations, and then turned back again with the stern
resolve that 'they would all take a share in the bloodshed’.
But meanwhile Labda, her suspicions aroused, had hidden the
baby in a box. And so Kypselos grew to manhood, overthrew
the Bakchidai, and became tyrant of Corinth . 238
It would be unwise to read too much into this silly story,
but the peculiar marriage custom must be accepted as a fact,
and this gives the key to the rest. Wade-Gery has suggested
that it ‘was perhaps due to their dislike that an heiress’s portion
should pass outside ’. 238 But the patriarchal law of the heiress
operates only in default of male heirs, not as a general rule.
We have seen in the present chapter from numerous examples
that the continuous intermarriage of near kin is a means of
obviating succession from mother to daughter in favour of
father and son. Assuming then that what we have here is a
normal instance of matriarchal endogamy, nothing further is
needed to explain why the men hesitated to kill the baby or
why the baby grew up with a claim on their inheritance. He
was their fellow clansman.
Lastly, it is worth noting that in a number of Aegean
islands, including Lesbos, Lemnos, Naxos and Kos, matrilineal
succession to real property was the rule at the end of the
eighteenth century a.d. The facts were reported by an English
traveller, John Hawkins, who wrote:
At the close of the year 1797 I transmitted to Mr. Guys as the result of
those enquiries which it had been in my power to make: that in a large
238 Hdt. 5. 92 y-e. 230 H. T. Wade-Gery in CAH 3. 534.
V
MATRIARCHAL PEOPLES OF THE iEGEAN 203
proportion of die islands of the Archipelago the eldest daughter takes as her
marriage portion the family house, together with its furniture, and one third
or a larger share of the maternal property, which in reality in most of these
islands constitutes the chief means of subsistence; that the other daughters,
as they marry off in succession, are likewise entided to the family house then
in occupation and the same share of whatever property remains; finally,
that these observations were applicable to the islands of Mytilin, Lemnos,
Scopclo, Skyros, Syra, Zea, Ipsera, Myconi, Paros, Naxia, Siphno, Santorini,
and Cos, where I have either collected my information in person or had
obtained it through others. 24 **
I am not in a position to explain this remarkable survival
or revival. That could only be done by embarking on the
unexplored subject of Greek land-tenure under the Byzantine
and Ottoman Empires. I mention it, because those scholars
who find it impossible to believe that anything so un-Greek
as matriarchy ever existed even in the prehistoric ^Egean may
be reassured to know that it was flourishing there in their
great-grandfathers’ time.
240 Hawkins in Walpole 392.
VI
THE MAKING OF A GODDESS
1. Childbirth and Menstruation
In the earliest phases of human society collective labour was
a condition of survival. Food-gathering and hunting, at a low
technical level, required many hands. There was no danger in
too many, because the surplus could always move away, but
too few meant death. Production of the means of subsistence
was inseparable from reproduction of the group itself. And if
the technique of production was precarious, so was that of
reproduction. The infant mortality of primitive peoples is
enormous. The magical rites that cluster everywhere round the
event of childbirth sprang from material necessity. 1
Similar conditions recur at a higher level with the discovery
of agriculture. So long as the new technique was rudimentary,
tremendous efforts were needed to make a clearing in the
forest and to keep it clear. The settlement was besieged with
unknown dangers — infectious diseases as well as wild beasts
injurious to health and wealth. In. the Germanic languages, as
in the Semitic, to till is to build (German bauen, Arabic
’ amara ). 2 It was necessary to tame the wilderness, which
fought back savagely. In such conditions it was impossible
for a single family to settle alone. Safety lay in numbers.
The crops, tended laboriously by the women, were blessed or
blighted by goddesses of childbirth.
The dread inspired by the magic surrounding the repro-
ductive functions of women was reinforced from the outset by a
powerful taboo. On this subject Briffault writes:
Although nothing exists in animal psychology exactly corresponding to a
taboo or formulated prohibition, there is one relation, and one only, in
1 1 do not mean that the rites were consciously designed for the survival
of the species, but simply that they were the ideological expression of
the maternal impulse.
2 Robertson Smith RS 95-6.
VI THE MAKING OF A GODDESS 205
which an interdict is normally imposed from without on the most potent
animal impulses. . . In the mammalian female sexual congress is not func-
tional or desirable during pregnancy and lactation, and the male is at these
times invariably repelled. In die human female, though the period of sexual
activity is more continuous than in many animals, a further interruption
appears in the form of definite menstruation. . . . The repulse of the male by
the female presents the analogue, and the only one, of a ‘prohibition’ among
animals. It can only be enforced among animals by the actual resistance or
escape of the female; it cannot therefore, in animal psychology, exactly cor-
respond to a formulated prohibition. Only traditional heredity can do this.
It is at the human level only, through the medium of language, that a pro-
hibition can acquire the status of a recognised principle. 3
As the first taboo, the ban on sexual intercourse during pregnancy
and menstruation became the prototype of all subsequent taboos.
It is important to observe that the magic of human fecundity
attaches to the process, not to the result — to the lochial dis-
charge, not to the child itself; and consequently all fluxes of
blood, menstrual as well as lochial, are treated alike as mani-
festations of the life-giving power inherent in the female sex.
In primitive thought menstruation is regarded, quite cor-
rectly, as a process of the same nature as childbirth . 4
This magic is ambivalent. Its very potency makes it some-
thing to be feared. It is a source of energy, like an electric
current, which without proper control can do a lot of damage.
So with the taboo. From one aspect the woman who may not
be approached is inviolable, holy; from another aspect she is
polluted, unclean. She is what the Romans called sacra, sacred
and accursed. And hence in patriarchal society, after woman
has lost her control of religion, it is the negative aspect that
prevails. Not only are her sexual functions treated as impure
in ' themselves, but the same condemnation attaches to her
feminine nature as such. She becomes the root of all evil, Eve,
a witch.®
These ideas are universal. There is no sphere of human life
in which a greater uniformity can be observed than in the
treatment of menstrual and puerperal women. The subject is
discussed at length by Briffault, who has collected examples
from every branch of the human race and every stage of culture.®
All that need be done here is to summarise the significant features
3 Briffault 2. 364. 4 lb. 2. 366. 6 lb. 2. 407. « lb. 2. 364-439.
VI
206 STUDIES IN ANCIENT GREEK SOCIETY
with a view to demonstrating their bearing on Greek religion.
Among the Herero tribes of South Africa the herdsmen
bring the morning’s milk every day to a woman in childbed,
who consecrates it with her lips.’ In North America, when the
corn is attacked by grubs, menstruating women go out at
night and walk naked through the fields . 8 Similar customs
still survive among the European peasantry . 8 Pliny recom-
mended as an antidote to noxious insects that menstruating
women should walk through the fields with bare feet, loose
hair, and skirts drawn up to the hips. Demokritos, according
to Columella, held the same opinion: the women, he said,
should run round the crop three times with bare feet and
flowing hair . 10 The idea was evidently to diffuse the fertile
energy with which the female body was believed at such times
to be charged. Elsewhere the energy is regarded as inherent
in their sex. Among the Zulus, for example, the girls who
perambulate must be naked but need not be actually men-
struating at the time . 11 This is the origin of the well-known
women’s rite of exposing the genitalia by drawing up the
skirts — a rite which in Greece was especially associated with
Demeter ; 12 and the custom common to many Greek cults of
female votaries walking in procession without shoes, head-
bands or girdles belongs to the same circle of ideas . 13 Eventu-
ally, purged of their superstitious dross, these girls, dancing
and singing through the fields unveiled, dishevelled, and un-
sandalled, became a traditional conceit, one of the prettiest in
Greek poetry . 14
7 Briffault 2. 410. 8 Schoolcraft 5. 70. 8 Briffault 2. 389, 410.
Plin. NH. 28. 78, Colum. HR. 11. 3. 64. 11 Krige 200.
12 Clem. Pr. 2. 20-1, Hdt. 2. 60. 2, D.S. 1. 85. The use of obscene
language — a constant feature of Demeter-worship — is a modification
of this custom: Horn. H. 2. 203-5, D.S. 5. 4, Luc. D. Mar. 7. 4 sch., Hdt.
5- 83. 3, Paus. 7. 27. 10, Hsch. ytipupfs, Suid. £k tOv &[ia§cov aKcbpiicrra,
Plu. M. 417c, A.R. 4. 1701-30, Apld. 1. 9. 26, 2. 5. xi, Thphr. HP.
7. 3. 3, 9. 8. 8; see further Briffault 3. 204.
13 Call. Crr. 1-6, 125-6, SIG. 736. 4, Nonn. D. 9. 243-8, Verg. A.
4. 509, Ov. F. 3. 257.
14 Call. Cer. 124-5, Bion 1. 21-2, Opp. Yen. 1. 497-8, Nonn. D. 5.
374, 405-7, 8. 16-19, 9. 248, 14. 382, Longus x. 4, 2. 23, Ach. Tat.
1. 1. 7, cf. Alcm. 1. 15, A. Pr. 140, S. OC. 348-9, Theoc. 19 (24) 36,
Polites E 74. 1 1 3-5.
VI THE MAKING OF A GODDESS 207
Faith in the life-giving properties of pregnancy and men-
struation has been everywhere combined, without any sense of
incongruity, with the deepest horror and aversion . 1 ■ A woman
in either condition must be segregated as strictly as one who
has touched a corpse. For a man even to set eyes on her may be
fatal to him. A mere touch of her hand or foot will maim the
cattle or blight the crops. In the case of childbirth access to the
hut in which she is confined is forbidden to all except the
midwives, and when the confinement is over, all clothes and
cooking utensils, together with the placenta, umbilical cord,
and all traces of blood, must be carefully destroyed or removed
to a place where there is no risk of anyone inadvertently
touching them or treading on them . 1 8 Among some peoples it is
sufficient to deposit them on a road or at a crossways, the idea
being that the pollution will be carried away by passing
travellers . 17 The woman herself must also submit to ablutions
and other purifications before she is readmitted to society.
During menstruation she is usually obliged to leave the village
altogether and retire into the forest, where she remains in an
isolation hut, alone or with other women in attendance. In
this case, too, before her return, she must obliterate all' vestiges
of her pollution and bathe in running water. A girl’s first period
is commonly the subject of special precautions, because it
coincides with her introduction to sexual life, and, where
women’s rites of initiation survive, it is included in them . 18
An account of the initiation of girls among the South African
Bantus will serve as an example.
When a girl feels her first period approaching she chooses from
a neighbouring village a married woman to be her ‘foster-mother’,
and when the day comes she runs away to this foster-mother
is Briffault z. 365-90.
18 Add to Briffault* s references Earthy 69-70, 75, Roscoe B (19x1)
21, 54, cf. BB 159: 'All the sweepings were thrown in some place where
they could not be trodden on or disturbed, and where future sweepings
from the house and all excrements from the child could also be thrown.'
17 Junod I. 200-1, cf. 2. 478; Petron. 134. In PI. Leg. 873b it is enacted
that the crossroads at which the bodies of murderers are to be thrown
must lie beyond the city boundaries, cf. 855a; see further Hastings r.v.
Crossroads.
18 Briffault 2. 371-2.
208 studies in ANCIENT GREEK-SOCIETY VI
‘to weep with her'. Her seclusion lasts a month. Three or
four girls usually take the. initiation together. They are shut
up in a hut, which they may not leave except with covered
faces. Every morning they go down to a pool to bathe, escorted
by initiated women who sing obscene songs and carry sticks
to drive away any man that may cross their path, because it is
believed that, if a man should see them, he would be struck
blind on the spot. After returning to the hut, though dripping
wet and shivering with cold, they are not allowed near the fire
but are scratched, teased, and tormented by the older women,
who keep up their bawdy songs, instruct them in sexual mat-
ters, and warn them never to reveal anything about the blood
of the menses to a man. At the end of the month the girl is
restored to her mother and entertained to a feast. She has
‘finished her misfortune ’. 1 9
In ancient Greece women in childbed were believed to be
under a pollution as grave as bloodshed or contact with a
corpse . 20 At Eleusis all those who had committed manslaughter,
or touched a corpse, or approached a woman in childbed, were
excluded from the Mysteries until they had been purified . 21 In
Hesiod’s Works and Days the men are warned not to wash
in water in which a woman has washed previously . 22 From
inscriptions we learn that women were not admitted to the
temples for so many days after menstruation or childbirth,
and then only after they had purified themselves by bathing . 23
In ancient Italy the horror excited by menstruation was as
great as it is among savages to-day. Pliny’s confidence in its
beneficial effects when properly applied did not weaken his
conviction that in general it was pernicious:
Hardly can there be found a thing more monstrous than is that flux and
course of women. For if during the time of their sickness they may happen
to approach or go over a vessel of wine, be it never so new, it will presently
10 Junod LSAT i. 177.
20 E. IT. 381-3, cf. Plu. M. 1 70b, Thphr.. Char. 16. 9, Ar. Lys. 912-3.
The Zulus treat the following categories as undean: pregnant and menstruating
women, mothers of infants, persons who have just had sexual intercourse,
attended a funeral, or touched a corpse (Krige 82).
=1 Porph. Ahst. 4. 16, Theo Sm. 14, cf. Apld. 2. 5. 12.
22 Hes. Op. 753.
20 SIC. 982-3, 1042.
VI THE MAKING OF A GODDESS 209
sour; if they touch any standing corn in a field, it will wither and come to no
good. Also, let them in this estate handle any grasses, they will die upon it;
the herbs and young buds in a garden, if they do but pass by, will catch a
blast and burn away to nothing.®*
These superstitions enable us to interpret a curious Sicilian
folktale:
Hera bore to Zeus a girl named Angelos, and Zeus gave her to the nymphs
to nurse. When she grew up, she stole the myrrh that Hera used to rouge
her face. When Hera found out, she was going to punish her, but the girl
ran away to the house of a woman who had just given birth and from there
to some men who were carrying out a corpse. Then Hera gave up the pursuit,
and Zeus told the Kabeiroi to take charge of the girl and cleanse her; so the
Kabeiroi took her away and cleansed her in the Lake of Acheron . 2 b
Angelos was identified with Artemis . 20 Why she stole her
mother’s rouge will appear in a moment, but it is already
plain that what prompted her flight from home and her contact
with the twin pollutions of birth and death was a third pollu-
tion of the same nature, which the storyteller has suppressed.
Aristotle, Pliny, and other naturalists, ancient and medieval,
believed that the embryo is formed from the blood retained in
the uterus after the stoppage of menstruation . 27 This is the
blood of life. Hence the commonest method of placing persons
or things under a taboo — menstrual, lochial, or any other inter-
dict formed on this original pattern — is to mark them with
blood or the colour of blood. And in keeping with tire am-
bivalent nature of the taboo itself this sign of blood has the
double effect of forbidding contact and imparting vital energy.
It is a worldwide custom for menstruating or pregnant women
to daub their bodies with red ochre, which serves at once to
warn the men away and to enhance their fertility. In many
marriage ceremonies the bride’s forehead is painted red — a
sign that she is forbidden to all men save her husband and a
guarantee that she will bear him children . 20 This is the origin
of cosmetics. Among the Valenge, a Bantu tribe, every woman
keeps a pot of red ochre, which is sacred to her sex and used to
paint her face and body for ceremonial purposes . 29 Of the
24 Plin. NH. 7. 64 tr. Holland. 20 Theoc. 2. 12 sch. 20 Hsch. 'Ayy&ov.
27 Arist. GA. 2. 4, PA. 2. 6. 1, Plin. NH. 7. 66, Briffault 2. 4, 44.
28 Briffault 2. 4x2-7.
29 Earthy 123, cf. 73, 76, Hollis NLF 58, Burkitt P 222-3.
O
210 STUDIES IN ANCIENT GREEK SOCIETY VI
many occasions for which she needs it the following may be
noted. At the end of her confinement both mother and child
are anointed with it: in this way the child will live and the
mother is restored to life. At initiation the girl is painted red
from head to foot: so she is born again and will be fruitful. At .
the conclusion of mourning, after stepping over a fire, the
widow is painted the same colour: so she returns from the
contamination of death.
Red is renewal of life. That is why the bones from upper
palaeolithic and neolithic interments are painted red. 30 The *
symbolism becomes quite clear when we find, as we commonly
do, that the skeleton has been laid in the contracted or uterine
posture (pp. 48, 55). Smeared with the colour of life, curled up
like a babe in the womb — what more could primitive man do
to ensure that the soul of the departed would be born again?
2. Moon-worship
It is a commonplace among all peoples of mankind that the
reproductive functions of women are regulated by the moon.
Whether there is any scientific foundation for the belief is
doubtful. Recent research is against it, though, since the data
are drawn from civilised women, the result is not conclusive. 31
Whatever the truth may be, the menstrual period coincides so
closely with the lunar that the idea was bound to arise of a
direct connection between them. This belief underlies all
primitive customs pertaining to the moon, which are so
deeply imbedded that even civilised peoples, who have long
discarded moon-worship/ retain them almost intact on the
lower levels of folklore and superstition.
In magic the moon receives far more attention than the
sun, and it is universally the first time-keeper. 02 The original
unit of the calendar was the lunar month of twenty-seven or
twenty-eight nights, which was divided in two by the full
moon. 33 Later a tripartite division was obtained by separating
so Burkitt P 163, 184, 191, Childe DEC Index s.v. Ochre.
3 1 Gunn 872, cf. H. M. Fox SSM 75, 80, LPR. 547.
32 Nilsson PT 148-9, cf. Briffault 2. 577-83.
33 Nilsson PT 155.
VI
THE MAKING OF A GODDESS
21 1
the periods before and after the full, 8 * and finally the Babylonians
rearranged the month in four quarters, which are the origin of
our weeks . 85 The dependence of the calendar on the moon is
reflected in the common base underlying our words for 'moon’,
'month', and 'measure '. 86
The moon is the cause of menstruation. In the Murray
Islands, to cite a typical instance, the moon is a young man
who ravishes the women and so causes their discharges of
blood . 37 Since the menstrual blood is believed to be the material
of the embryo, the discharge is regarded as a form of abortion
— what is still popularly known as a 'moon-calf'. From these
premisses it should follow that impregnation too is caused by
the moon, and in primitive thought it does follow. The moon,
say the Maoris, is the real husband of all women . 38 The truth
34 lb. 167-70. 35 Ih. 1 71, Langdon BM 86-7.
36 The old view that IE *mater is connected with this base has been
abandoned, ma- being now explained as a Lallwort (Walde-Pokorny s.v.)
but the two interpretations are quite compatible; for 'the evolution of
language is a process of differentiation' (Br£al 33).
37 Briffault 2. 583-4. 38 lb. 2. 432.
VI
212 STUDIES IN ANCIENT GREEK SOCIETY
about procreation is only recognised when paternity has ac-
quired a social value, and even then the older belief lingers on.
Women conceive from the moon. Accordingly in primitive
languages the moon is normally spoken of as a male, the Lord of
the Women. 80 It is still masculine in Slavonic and Germanic,
and it was once so in Celtic, Greek and Latin. The Greek
selene is a substitute for mine, a feminine derivative of min,
which survived as the word for ‘month*. In Anatolia the moon
continued to be worshipped as a male, whom the Greeks
called Men. 40 From the idea that women conceive from the
moon arises the further notion that through him they become
inspired or possessed. Hence the traditional associations of the
moon with hysteria, epilepsy, and all diseases regarded as
divine. 41 The connection between lunar influence and lunacy
speaks for itself.
Emanating as it did from the sexual life of women, moon-
worship became involved with their social functions. It was
their task to draw water, tend the plants, secure plenty of dew
and rain. The moon was accordingly regarded as the cause of
growth in vegetation, the source of all lifegiving waters. In
India it is ‘the bearer of seed, the bearer of plants’; in Baby-
lonia it was the fountain-head of all plant life. Hence its
identification with a sacred plant or tree, like the Indian soma
or North American maize. Natives of the Rio Grande used to
say, ‘While the moon is growing, the sap is always flowing’. 48
Vegetable juices, especially aromatic gums used for incense and
unguents, derive their virtues from the moon. ‘The virtue of
the gum acacia as an amulet’, writes Robertson Smith of the
Semites, ‘is connected with the idea that it is a clot of men-
struous blood, i.e. that the tree is a woman.’ 48
The Egyptian moon-goddess Nit was the inventor of the
loom, and in European folklore the moon is still a spinner.
These too are women’s tasks, and in many primitive traditions
we find the moon engaged in grinding corn, making pots, or
cooking. 44 Further, since magic had once been controlled by
80 Nilsson PT z. 583-97. *o Roscher LGRM 2. 2687.
41 Briffault 2. 608-10.
i2 lb. 2. 624-38. Cf. A. Jercmias in Roscher LGRM 4. 1470.
43 Robertson Smith RS 133. ** Briffault 2. 624-8.
VI THE MAKING OF A GODDESS 213
women, it is a potent source of charms and spells, especially
for love-making; and when feminine magic has been banned,
it remains, as it still- is, patron of the black art of witch-
craft. 46 -
The idea of the moon as the source of fertility was expanded
by the subjective interpretation of actual lunar phenomena.
The moon waxes and wanes, grows and decays, dies and is
born again. It becomes a universal symbol for the renewal of
life. Among all primitive peoples the belief in resurrection after
death is associated with the rising moon. 40 In this way it was
brought into relation with other symbols of the same sort,
especially the serpent, whose significance has been explained
in Chapter IV (p. 1 19). In Australia and Melanesia the moon is
said to cast its slough month by month, and conversely the
serpent is everywhere a seducer of women and guardian of
sacred waters. 47 This association of ideas is assisted by the
animal’s phallic shape and its habit of frequenting pools and
springs. 48 The place of the serpent in fertility ritual may be
illustrated from the secret society of the Mpongwe women in
West Africa. Their meetings, held in the depths of the forest,
are so secret that, though they have a recognised head, called
the Mother, nobody knows who she is. Every woman must
catch one of the little snakes that live in the mangrove roots;
then they strip naked and with the snakes in their hands strike
up lascivious songs, singing and dancing all through the night
till they drop from exhaustion. 49
The moon’s lifegiving virtue is also believed to reside in
stones, especially crystals and translucent gems, and in human
bones and hair. 60 The importance attached to tooth-evulsion
and hair-cutting at initiation (pp. 47-8) arises from the fact that
these parts of the body possess the property of self-renewal.
Where sacrifices are offered to the moon, there is a remarkable
uniformity in the choice of victims. 61 These are principally the
hare, the goat, the pig, all of which are still prominent in
witchcraft; the dove, especially characteristic of Semitic
women’s cults; and the cat, which owes its proverbial nine
46 lb. 2. 620-3. 40 lb. 2. 651-2. 47 lb. 2. 664-73.
48 lb. 2. 66 7, Roscoe BB (1923) 43 -4. 40 Briffault 2. 548.
60 lb. 2. 692-4, 702-9. 61 lb. 2. 610-23.
VI
214 STUDIES IN ANCIENT GREEK SOCIETY
lives to this association.-In ancient Egypt the cat was an object
of worship, and Plutarch explains why:
The variegated colouring of this animal, its nocturnal habits, and its
peculiar manner of reproduction combine to make it a fitting symbol for the
moon. It is said to produce one kitten at the first birth, two at the second,
and one more every time until the seventh, making twenty-eight kittens
in all, which is the number of days in the month; and though this may
savour of the fabulous, there is no doubt that a cat’s eyes appear to grow
larger and more luminous at the full moon, smaller and duller as it wanes. 62
As BrifFault remarks, ‘the rule that the sacrificial animals of
women belong to small species is universal’. His explanation is
that ‘women being in most cases unable to offer large animals
such as are used in sacrifice by hunters and herdsmen, are
generally confined in their choice to smaller species’. 63 This is
true of the later stages, but the initial factor was more pro-
bably the domestication of animals. It is believed that this
began with the huntsmen bringing home the young of smaller
species, which the women kept as pets. 64
Lunar magic has also had an influence on dancing. The
Iroquois dance in honour of the moon for the sake of its health
when it is sick, the Californians to prevent it waning. The
Dieguenos of southern California used to run footraces
regularly at the new moon to help it grow, and the Pawnees
assert that their ball dances were instituted by the Great
Hare in memory of his brother the Wolf, who was the waning
moon. 66 In such cases the ball is a mimetic symbol.
3. The Moon in Popular Greek Religion
Aristotle’s theory of the formation of the embryo from the
menses is in harmony with the belief, which he shared with
Empedokles, that menstruation occurs normally towards the
end of the month, when the moon is on the wane. 66 His ac-
count of the spinal marrow follows the same lines. ‘The parts
of the body’, he says, ‘are formed from the blood; the embryo
62 Plu. M. 376s* 63 Briffault 2. 619.
64 Thurnwald 77, Frazer TE 1. 14-5. ss Briffault 2. 746-9.
66 Arist. HA. 7. 2 (582b), GA. 2. 4. 9, Sor. Gyn. 21 (Rose 185), cf. Gal.
9. 903; Roscoe B (191 1) 24.
VI ' THE MAKING OF A GODDESS 215
is fed on blood; and in die same way the marrow represents
the element of blood in the bones .’ 67 The word aiStt, marrow ,
was extended to the idea of eternity. A direct connection be-
tween the marrow and the moon is not attested, but something
of the sort must have been in yEschylus’s mind when he
described the marrow as ‘reigning’ within the breast, because
this term was used in astrology for the influence of celestial
bodies . 68 The moon figures as a symbol of eternity in the myth
of Endymion, who during his everlasting sleep was visited
nightly by Selene . 68
In Greece, as elsewhere, the moon was popularly regarded as
the source of fertilising moisture. The dew, according to
Plutarch, is heaviest at full moon, and in poetry it is a daughter
of the moon . 00 In Stoic doctrine the moon draws its sustenance
from springs and streams . 01 The moon, says Pliny, brings forth
the moisture which the sun consumes; and, according to
Cicero, it releases a flow of moisture which fosters the growth of
living creatures and brings to maturity everything that rises
from the earth . 08 Epilepsy and allied disorders were ascribed to
the same influence , 08 and the symptoms of other maladies were
thought to rise or subside according to the lunar phases . 04 The
full moon was the best time for sowing or planting , 06 and for
67 Arist. PA. 2. 6. i. 68 G. Thomson AO 2. 13.
69 Apld. i. 7. 5, Paus. 5. 1. 4.
00 Plu. M. 36yd, 659b, Alcm. fir. 43, cf. Thphr. CP. 4. 14. 3, Macr.
Sat. 7. 16. 31, Gal. 9. 903, Horn. H. 32. 11-2, Verg. C. 3. 337, Nonn.
D. 40. 376, 44. 221, II. 23. 597-91 A. fr. 44, A. 1390-1, A. R. 3. 1019-21.
01 Porph. Ant. n, cf. Plu. M. 659b, Plin. NH. 2. 223, Arist. Mete. 1. 10.
02 Plin. NH. 20. 1, Cic. ND. 2. 19. 50.
03 Gal. 9. 903, Macr. Sat. 1. 17. 11, Artem. 104. 14, Orph. L. 50, 474-84,
cf. Ar. Nu. 397 sch., Hsch. Pekkeo&tivos, Psalms 121. 6; Roscoe B (1911) 24,
Earthy 73.
64 Gal. 19. 188, Plin. NH. 28. 44. So with prophecy: at Argos there ■
was a shrine of Apollo at which monthly oracles were delivered by a
woman after drinking ram’s blood: Paus. 2. 24. 1. The prophetic trance is
probably the 'epileptic equivalent’ ( Dammerzustani ), a condition similar
to the epileptic fit but without loss of consciousness or violent physical
convulsions (Bleuler 338). It is possible that Gk. uccvtiki'i and nw(« both go
back to the same root as jify (see above n. 36).
06 Pall. 1. 6. 12, Gp. 1. 6. 1, 5. 10. 1, Lyd. 2. 8.
VI
216 STUDIES IN ANCIENT GREEK SOCIETY
weddings . 68 New-born infants were taken out of doors by:
their nurses and' ‘shown to the moon ’. 67 One of the most
efficacious amulets was the moonstone, which was strung into
children's necklaces, hung on fruit-trees, and used as a love
charm and a cure for epilepsy . 08 On the same principle the -
time for cutting plants, felling, and sheep-shearing was when
the moon was waning 68 — a rule scrupulously observed by the
Emperor Tiberius in having his hair cut . 70 Neglect of this
precaution resulted in baldness.
The connection with the serpent appears in the popular
belief, accepted by Aristotle without demur, that a snake has as
many ribs as there are days in the month . 71 As a guardian of
waters the snake appears in many Greek myths, the best-
known being the story of how Apollo conquered Delphi:
Near at hand was a bubbling spring, and there with a shaft from his
strong bow Apollo slew the dragon, a fearful overgrown monster, which had
done untold harm to men.? 2 —
The serpent was equally familiar as a ravisher of women, as we
see from the inscriptions discovered in the temple of Asklepios,
the snake hero, at Epidauros. On one occasion a woman visited
the temple to cure herself of sterility, spending the night there
in accordance with the usual practice. She dreamt that the god
came to her with a snake, with which she had sexual congress,
and nine months later she was delivered of male twins. 73 ^Here
the snake is the god, who impregnates his votary through his
animal medium. It may be added that the healing powers of
the snakes kept in the temple of Asklepios were attributed to
GB E. IA. 716-7, Pi. L 8. 44-5, D. Chr. Or. 7. 245R. Parturition was
easiest at full moon: Cic. ND, 2. 46. 210, Plu. M. z 8 zd, 6 $ 8 f, 93 of, II.
21. 483 sch A.
07 p lu. M. 6 $ 8 £, cf. Ath, 139a, D.S. 5. 73; Briffault 2. 590-1, cf. Nilsson
PT 149-54.
08 Dsc. 5. 159, Hsch. oeXrivfs, cf. Gal. 9. 859, Azl. NA. 14. 27, Plaut. Ipid.
5. I. 33, Plu. M. 282a, 287f-288b.
flD Plin. NH. 18. 321-2, Varr. RR. 1. 37.
76 Plin. NH. 1 6. 194.
71 Arist. HA. 2. 17. 23, Plin. NH. 11. 82.
72 Horn H. 3. 300-3, cf. Apld. 3. 6. 4, Paus. 9. 10. 5.
73 SIG. 1169. 19, cf. 1168. 1 12. On the snake affinities of this god see
Frazer PDG 3. 65-6.
VI
THE MAKING OF A GODDESS
217
the animal’s faculty of renewing its life by changing its skin. 7 *
In further illustration of this circle of ideas it wifi suffice for
the present to provide the West African snake dance described
above (p. 213) with a parallel from Macedonia:
From early times the women of this country have been addicted to Orphic
v and Dionysiac orgies. . . . Surpassing the others in her zeal, Olympias made the
HG. 9. A Mttnai: Attic vase
cites still more barbarous and abandoned by carrying in the dances huge tame
snakes, which kept creeping out of the ivy in the mystic cradles and coiling
round the women's wands and crowns — a sight that struck terror into the men. 7 ®
Olympias was the mother of Alexander the Great. It is related
that some time before he was bom her husband came home
and found her asleep in bed with a snake beside her. 76
7* Ar. PL 733 sch. 76 plu. Alex. 2. 76 P] u . Alex, z, cf. Suet. Oct. 94.
- /
2l8 STUDIES IN ANCIENT GREEK SOCIETY
VI
4. Herbal Magic
We are now in a position to uncover the primitive magic
that lies beneath the worship of the stately goddesses of the
Greek pantheon.
Herbal magic is everywhere die province of women. In her
study of the Valenge Miss Earthy writes:
Nearly all trees and plants have a magical value. If the women saw me
gathering botanical specimens, their curiosity was at once aroused, because
plants are associated in their minds with recipes for magic or medicine.’ 7
Ancient Greek herbal lore, which can be studied in the pages
of Dioskorides and Pliny, has not received as much attention as
it deserves from students of Greek religion.
The root of the peony, which was called menion or selenogonon ,
implying that its virtue was derived from the moon, was ad-
ministered at menstruation and childbirth. 78 The dittany
(diktamnos) was used to assist parturition and woven into
chaplets for Eileithyia, the goddess of childbirth. 79 The myrtle
prevented premature delivery by closing the uterus; the lily
was a check to menstruation. 80 These flowers were sacred to
Aphrodite. 81 The lygos, a species of withy, was variously believed
to induce or arrest menstruation. 88 At Sparta Artemis was
called Lygodesma, ‘withy-bound’, because her image was said
to have been discovered in a bed of withies, 83 and the statue
of the same goddess at Agra (Attica) was decorated with gar-
lands of withy. 84 - In the same way the Hera of Samos was saicl
to have been born under the withy-tree that grew in her
sanctuary. 86 The galingale (kypeiros) was made into potions for
77 Earthy 24.
78 Dsc. 3. 157, Plin. NH. 26. 151. On the use of herbal medicines for
menstruation in medieval and modem Europe see McKenzie 284-6.
79 Thphr. HP. 9. 16. i, Arat. 33 sch.
89 Plin. NH. 23. 159-60, 24. 50, 2i. 126. The Myoj or &yvos (Plin. NH.
24.- 59) is the Vitex agnus castiu : Hort 2. 437.
81 Paus. 6. 24. 7, Ov. Met. 10. 512. The pomegranate was consecrated
to her in Cyprus: Ath. 84c. At Boiai (Laconia) the myrtle was sacred to
Artemis Soteira, ‘saviour’ of women in childbirth (p. 275): Paus. 3. 22. 12.
83 Dsc. 1. 134, Plin. NH. 24. 59-60. as p a ns. 3. 16. n.
84 E. Hip. 73 sch. (read Myw for tayco). se Paus. 7. 4. 4.
-VI
THE MAKING OF A GODDESS
219
opening the womb; 80 the helichryse assisted the menses, and
was used by Spartan girls to make crowns for Hera. 87 In
Alkman’s hymn to Hera a girl prays, 'I beseech thee and bring
unto thee a wreath of helichryse and galingale/ 88 One of the
functions of Juno, the Roman Hera, was to assist menstrua-
tion, 88 and Hera, Artemis, Aphrodite were all worshipped as
goddesses of childbirth. 80
Of all the plants in Greek lore the most familiar was the
pomegranate. Demeter was constantly portrayed
as holding in her hand a poppy or a pomegranate
or both. 81 The image of Victory Athena at
Athens had a helmet in the right hand and a
pomegranate in the left. 88 At Olympia there
was a statue of the athlete Milon holding a
pomegranate. Milon was a priest of Hera. 83
Hera’s statue at Argos had a sceptre in one
hand and a pomegranate in the other. In re-
ferring to it Pausanias remarks: ‘I will say
no more about the pomegranate, because the
story connected with it is in the nature of a
secret.' 84 What was the secret?
The fruit of the pomegranate is a brilliant red.
So is the seed (kokkos) which by yielding a
common dye gave Greek its word for scarlet
(k6kkinos).*s The pomegranate was a sign of blood.
3 ® Plin. NH. zi. 1 18. The seeds were eaten roasted to
arrest menstruation. The KCmsipoi is the Cyperus longus :
Hort2. 461.
87 Plin. NH. 21 . 148, Ath. 678a. 88 Alcm. 24.
88 Varr. ap. Aug. CD. 7. 2.
8 « Famell CGS 1. 196, 2. 444, 655-6.
81 Roscher LGRM 2. 1342-3, 1345.
88 Heliod. fr. z-FHG. 4. 425, cf. Farneil CGS 1. 327.
83 Paus. 6. 14. 6, Phiiost. Ap. Ty. 4. 28.
84 Paus. 2. 17. 4. Hera holds a basket of pomegranates in statuettes
from the Heraion at the mouth of the Sele in Lucania: Zanotti-Bianco 244;
see further G. W. Elderkin 429-31, Bossert 327.
88 Str. 630. There are two varieties, red and white, and it was pre-
sumably the latter that was used to check menstruation, on the principle,
still observed in the Balkans, that red flowers always stimulate the blood:
Kemp 37; see further McKenzie 247-9.
FIG. 10. Goddess
with pomegranate:
Attic statue
'VI
220 STUDIES IN ANCIENT GREEK SOCIETY
That is generally understood, but the real meaning of the sign
has been missed. It has .usually been regarded as a symbol of
violent death. 90 That is undoubtedly what it means in partic-
ular cases. It was said to have sprouted from the blood of
Dionysus when he was slain by the Titans; 07 it bloomed over
the body of the suicide Menoikeus, and was planted on the
grave of Eteokles by the Erinyes who had caused his death. 98
To dream of pomegranates portended wounds. 99 But these
applications are secondary. The pomegranate was used medi-
cinally for menstruation and pregnancy, 100 and this shows that
in the hands of Demeter it had the same value as the poppy,
which is expressly described as a symbol of fecundity. 101 In the
Eleusinian Mysteries and the Arcadian Mysteries of Despoina
(Persephone) it was taboo, 102 in allusion to a well-known in-
cident in the story of Persephone, which will be examined
presently. At Athens the women who kept the Thesmophoria
were required to abstain from pomegranates and from sexual
intercourse. Each night they slept on beds of withy, which had
the double virtue of checking the sexual impulse and scaring
away snakes. 193 Since they used the withy as an antidote to
sexual activity, we must suppose that they avoided the pome-
granate because it was a stimulant. Its colour was not primarily
the blood of battle but the blood of fertility — menstrual and.
lochial blood.
5. The Thesmophoria and Arrhephoria
The purpose of the Thesmophoria was to fertilise the crops.
Why then did the women take care to avoid fertilising in-
fluences? The answer to this question will show what happens
to primitive ritual when its original function is forgotten. 104
99 Frazer PDG 3. 184-5. 87 clem. Pr. 2. 16.
98 Paus. 9. 25. 1, Philost. Im. 2. 29. 4. 99 Artem. 1. 73.
199 Plin. NH. 23. 107, 1 12.
101 Eus. PE. 3. 11. 6, cf. Call Crr. 45, Theoc. 7. 155-7; A. J. Evans
PM 3. 458. The myrtle, which is allied to the pomegranate, was made into
crowns for the Eleusinian hierophants: S. OC. 683 sch., cf. Ar. Ra. 330 sch.
102 Porph, Abst. 4. 16, Paus. 8. 37. 7.
103 Clem. Pr. 2. 16, Plin. NH. 24. 59, Ael. NA. 9. 26, cf. Hsch. tcvtopov,
Ov. Met. 10. 431-5.
194 Theprincipal sources are Luc. DMcr. 2. 1 sch., 7. 4sch; Deubner 43-66.
VI THE MAKING OF A GODDESS 221
The festival took place towards the end of October. Some
months before — probably at the Skirophoria, held in June or
July — the women had sacrificed a number of pigs and deposited
them in a cavern as an offering to Demeter. Now, having kept
themselves in a state of purity for three days , 108 they entered the
cavern, clapping their hands to scare the snakes, and recovered
the decomposed remains, which were then mixed with the
seed-corn for the autumn sowing. The pig$ are described as
‘symbols of the birth of man and crops '. 106 A story was told that,
when Persephone was carried off into the underworld, a herds-
man who happened to be in the vicinity was engulfed together
with his swine . 101
Like the hare, dove, and other women’s animals, the pig
was believed to be exceptionally prolific . 108 It was therefore an
appropriate symbol of fertility. But it was more than that. It
was a substitute for the woman herself. That explains why the
106 The same rule was observed at the same festival at Abdera (D.L. g.
43) and Sparta (Hsch, Tprfjpspos) and in the cult of Isis at Tithoreia
(Paus. 10. 32. 14).
100 Luc. DMcr. 2. x sch. The principle is that both people and crops are
made to increase through the fertilisation (initiation) of women (D. A.
Talbot 86): hence such doublets as Damia and Au*esia (Hdt. 5. 82-3),
Hegemone and Karpo (Paus. 9. 35. 2), Dionysos Polites and Auxites
(Paus. 8. 26. x).
107 Clem. Pr. 2. 17.
108 Luc. DMer. 2 . 1 . sch.
VI
222 STUDIES IN ANCIENT GREEK SOCIETY
word for this animal (choiros) was used in vulgar language
for the pudenda muliebria . 10B The word kSkkos was used in
the same sense .* 10 The pig’s blood was a surrogate for the
woman's hathdrmata — menstrual and lochial blood. The. starting-
point of the festival was the primitive practice of secretly dis-
posing of these kathdmata, which were used to fertilise the
seed-corn . 111 Afterwards, when this function had been transferred
to the pig’s blood, the sexual activity of the women, deprived of
its positive value, was treated as a pollution and banned.
This type of ritual was not confined to Demeter. On a
certain night in the year two girls consecrated to Athena Polias
used to descend the Acropolis carrying something secret in a
box. At the foot of the hill was a cave. They went in, deposited
their charge, and returned with another object which had been
laid there on a previous occasion . 118 What the box contained is
not stated, but it can be inferred from the story of Erich-
thonios. Born in the form of a snake, Athena put him in a box,
which she gave to the daughters of Kekrops with strict instruc-
tions not to open it. They did, with the result that they went
mad and threw themselves from the Acropolis . 118 The snake
has the same meaning as the pig f , and the taboo on it has
developed in the same way.
This ceremony was known as the Arrhephoria, which means,
as Deubner has shown, the ‘conveyance of secrets ’. 114 Similarly,
ion Van:, jyj. z . 4. 10. no Hsch. j.v.
111 That the Greeks took the same precautions as other peoples (see above
n. 16) in disposing of the KaOdcppcrra, especially of women and epileptics,
is clear from Paus. 2. 31. 8, 5. 5. xo, 8. 41. 2, Hp. Morb. Sac. 4, and an
allusion to their original nature is perhaps embodied in the name of the
associated festival, the Skirophoria: Plin. NH. 7. 63, Arist. GA. 4. 7. 1,
Ath, 647a. The popular etymology (Ar. Ec. 18 sch.), referring to a white
umbrella (ok!pov=oki6Beiov), may be dismissed as a polite invention.
112 Paus. 1. 27. 2-3.
113 Apld. 3. 14, 6: fig. 36. All such stories of insanity caused by the sight of
sacred objects (Paus. 3. x 6. 9. 7. 19. 7, Derc. 7) go back to pre-deistic magic.
114 Deubner 1 12 <5tppri96pia= dpptyrcxpdpia (EM. &ppr)q>6poi), cf. -rfrpaxiios
(rfrpfiSpaxuos), Kidxpavov (kiov6kpccvov), &Kp 60 rrov (&Kpov 60 eTov). The alternative
form, fppTi<p6pia (Hsch. &ppr)<p6pot, IG. 2. 1397-85, 3. 902, 9x6) is due to
confusion with the cult of Pandrosos (epcrn = SpBaoj): see p. 261. The
Arrhephoria fell in the month of Skirophorion (EM. s.v.) and is probably
only another form of the Skirophoria.
VI THE MAKING OF A GODDESS 223
the Thesmophoria is the ‘conveyance of thesmoi’. A thesmos is
something ‘laid down’, normally a ‘law’ or ‘ordinance’, and in
later times the name was referred to Demeter in her capacity
of marriage goddess. 113 But it is obvious that such ritual is
older than marriage laws, and thesmSs was also used in a concrete
sense of things ‘stored’ or ‘deposited’ 1 16 — in this case, originally,
the kathdrmata. The two festivals have a common source in
women’s fertility magic.
6. Rites oj Ablution
Whereas Demeter was mainly concerned with the cultivation
of cereals, Artemis was a goddess of woodland, marsh, and
meadow. She was
the archer goddess who roams the hills, Taygetos or Erymanthos, delighting
in boars and stags, at play with the nymphs of the wild, while her mother
rejoices to see how beautiful she is, head and shoulders above the others —
all are lovely, but she is the first to catch the eye. 11 ’
At Letrinoi, on the banks of the Alpheios, there was a shrine
of Artemis Alpheaia. Once upon a time Alpheios fell in love
with her. He knew that, being pledged to chastity, she would
never yield to his advances, so he approached by stealth while
she was out with her nymphs at an all-night festival. But
Artemis and her companions had got wind of his intentions, so
they plastered their feces with mud, which made it impossible
113 Deubner 44-5. 116 Anacr. 58. 117 Od. 6. 102-8.
VI
224 STUDIES IN ANCIENT GREEK SOCIETY
for him to distinguish one from another . 118 This myth pre-
supposes a secret rite in which marriageable girls went down
to the river and assimilated the life-giving waters by smearing
themselves with slime. That medicinal properties were ascribed
to the Alpheios is shown by its name, the River of Leprosy
(alphas ). 119 The whole valley was rich in cures of this kind.
Further down the coast from Letrinoi was Lepreos, Lepers’
Town, with a shrine of Zeus Leukaios, god ‘of the white
sickness’, and there was a stream in the same district called
the Anigros, with a cave to which lepers made pilgrimages . 120
Other local myths follow the same pattern. Leukippos of
Pisa fell in love with Daphne. Disguising himself in women’s
clothes, he became her intimate friend by accompanying her
and her companions on their hunting expeditions. One day,
however, the girls decided to bathe in the Ladon. When
Leukippos refused, they stripped him forcibly and, discovering
his sex, stabbed him to death . 121 The Ladon is a tributary of
the Alpheios, rising on the slopes of Erymanthos. Another,
example comes from Boeotia. On a hot summer’s day, tired out
after hunting, Artemis found herself in a densely-wooded
valley, and was bathing in a spring called the Maiden’s Well
(Parthenia) when, attracted to the same spot with his hounds,
a man, Aktaion, set eyes on her. To prevent him revealing
what he had seen, the goddess turned him into a hind, which
the hounds immediately devoured . 122
All over the Greek countryside there were springs and
streams called Parthenia or Parthenios . 128 The girls used to
bathe at such places to purify themselves before religious
festivities , 124 and it was customary for brides to bathe before
118 Paus. 6. 22. 9, cf. Telesill. i. i» Str, 347, Lycoph. 1050-3 sch.
120 Paus. 5. 5. 5, 5. 5. 11, Str. 346. 121 p aU s. 8. 20. 3.
122 Hyg. F. 181, So on Mount Pholoe (Arcadia) Artemis shot Bouphagos
for attempting to rape her: Paus. 8. 27. 17.
123 Paus. 6. 21. 7, Str. 357, 457, Horn. H. 2. 99 etc.
124 Plu. Af. 77 if, 772b 1 cf* Ar. Ly. 913. Hence local myths of the type
recorded from Haliartos, where the infant Dionysus was said to have
been washed by his nurses in the spring of Kissousa: Plu. Lys, 28, cf. Paus.
9. 20. 4, and see further my A A 144* In Messenia, the KaO&ptiara from
the birth of Zeus were thrown by the nymphs into the River Neda, where
the children of Phigalia (probably when they came of age) used to dedicate
their hair: Paus. 8. 41. 2, cf. 8. 28. 2, 1. 43. 4, Poll. 3. 38.
VI THE MAKING OF A GODDESS 225
HG. 13. Artemis and Aktaion: Attic vase
the wedding in the local river or in water brought from the
river . 126 All such practices go back to initiation . 128 The act of
immersion served originally to purify the girls at their first
menstruation and at the same time to make them fruitful. The
nuptial water, which was described as ‘life-giving’, was
believed to have this effect. 12 ’ Brides of the Troad were even
more explicit. When they bathed in the Skamandros, they
used to pray to the river-god and say, ‘Skamandros, take my
virginity !’ 12 8 This shows that in Greece, as in other countries, it
had once been believed that the girl was actually impregnated
by the water. And it is fairly plain-that these girls who having
bathed become brides are the human prototypes of the
126 Th. z. 15. 5, Poll. 3* 43 > E. Pi. 344-8, Harp. XovTpc^ipos, Plu.
M.772 b.
128 Earthy 167: 'All washing of the body is more or less of a ceremonial
character. ... It is a type of cleansing from the moral defilement of illness,
death, and loss. . . .When people are ill they will not wash themselves
until they are well again.' The Dardaneis (Illyria) used to wash only twice
in their lives, at birth, marriage, and burial: Nic. Dam. 1 10.
127 Nonn. D. 3. 89, E. Pb. 347 sch.
128 Ps. iEschin. Ep. 10. 3.
P
VI
226 STUDIES IN ANCIENT GREEK SOCIETY
nymphs — the ‘brides’ (nymphai) who embrace the river-gods
and bear heroic sons.
7. The Daughters of Proitos
The Anigros, to which lepers resorted, had been known in
early times as the Minyeios. It was the same stream in which
Melampous had purified the daughters of Proitos and so healed
them of their madness (p. 196). We can now see a litde
further into the ritual behind this myth. 188
The girls went mad ‘when they reached maturity' 180 — that is
to say, at their first menstruation. In one version they are
driven mad by Dionysus, in another by Hera. Dionysus came
from the north, and according to Herodotus he was introduced
by Melampous himself. 131 This implies that the Peloponnesian
Agriania had been superimposed on an older cult of the same
nature. In the second version the girls are driven mad because
they have stolen Hera’s gold. 138 We remember how Angelos
stole her rouge (p. 209). And they were turned into cows. 1 38
They were restored, as I have said, in the Anigros or Minyeios, -
and after purifying them Melampous threw the katbarmata
into the stream. 184 In another version he purified them
near Lousoi at a shrine of Artemis Lousia, the Bathing
Artemis. 136
What was their disease? (The only difficulty in answering
this question arises from the fact that in primitive medicine
maladies now known to be distinct are confused. Leprosy is
not the same as epilepsy, but in primitive thought they have
a great deal in common. The leper was expelled because of his
contagion; the epileptic, on the approach of a seizure, ran out
of doors and away into the wilds. 136 Both became outcasts.
Hence, when ABschylus describes the physical condition of a
S r U tf r P erse f ute< ^ b y die Erinyes, he combines the symptoms
of the two diseases. 137 Again, epilepsy was regarded as possession
188 This section is based largely on Roscher SV 70-1.
130 Apld. 2. 2. 2, Hes. fir. 27=Str. 370. 131 Hdt. 2. 49.
182 Serv ‘ ad Ver 2 - E - 6 - 48 . 138 Prob. ad Verg. E. 6. 48.'
P aus . 5. 5. 10. xas p aus . 8 . r8. 8. cf. 2. 7 .8.
136 Hp. Morb. Sac. 4. 137 a. C. 277-95.
VI
THE MAKING OF A GODDESS
227
by an animal . 138 That explains why the Proitides were turned
into cows. Such therianthropic delusions, properly symptoms
of schizophrenia, are rare among civilised peoples, but among
savages still dominated by a totemic psychology they are
common . 139 Further, the Greeks believed that in women
epilepsy was caused by a stoppage of the womb, which pre-
vented menstruation . 140 They were of course mistaken. Such a
condition might lead to hysteria, but not to epilepsy. But
owing to the similarity of the symptoms these disorders were
identified, and long after Hippokrates had laid the foundations
of scientific medicine the belief persisted that hysteria was an
affection peculiar to women, caused by a stoppage of the
womb ( hystfra ). 141
The idea that a number of women should suddenly lose their
senses and rush out into the open country seems fantastic to
our minds, but scepticism is soon dispelled by the study of
primitive psychology . 142 It was not fantastic to the Greeks.
-Cases of mass hysteria are recorded from Sparta and Lokroi
Epizephyrioi, and in both the victims were women . 143 At
Sparta they were cured by the medicine-man Bakis under in-
structions from the Delphic Oracle. At Lokroi they would be
sitting quietly at their meal, when suddenly, as though in answer
to a supernatural voice, they would leap up in a frenzy and run
out of the town. They were cured by singing paeans to Apollo.
There are still some elements in the Agriania that we have
not accounted for — the infanticide, the pursuit, and the
slaughter of the hindmost. These must be held over till a later
stage of our enquiry . 144 But the general purport of the Argive
138 Hp. Mori. Sac. 4: hence its name Selcc viros (Aret. SD. 1. 4) the ‘divine
disease’; Junod LSAT 2. 479.
439 Roscher SV 71, Frazer PDG 5. 381-3, Bleuler 105.
wo Gal. xx. 165.
141 Arist. GA. 4. 7. 6.
142 Frazer A 147 quotes from I. H. N. Evans a case from Malaya: 'A
curious complaint was made in my presence by a Jakun man . . . that all
the women of his settlement were frequently seized by a kind of madness —
presumably some form of hysteria — and ran off singing into the jungle, each
woman by herself, and stopped there for several days and nights, finally
returning almost naked, or with their clothes tom to shreds. . . . They were
started by one of the women, whereupon all the others followed suit.’
W3 Ar. Av. 962SCI1., Aristox. 36. 144 Secmy AA 144-5.
228 STUDIES IN ANCIENT GREEK SOCIETY VI
myth is now clear. It is a projection of the terror inspired by the
magic inherent in the physiological processes of women.
8. Creek Goddesses and the Moon
It is now half a century since Roscher, working on a thorough
collection of the data, argued that virtually all the Greek god-
desses were in the first instance lunar deities. 146 His views
are not generally accepted. The attitude of his opponents,
led by Farnell, is that the evidence on which he relied is
drawn largely from Hellenistic sources, belonging to a time
when Greek religion had been thrown open to oriental
influences. The controversy is not settled yet, and never
will be, until students of Greek religion adopt a scientific
method.
Not all Roscher's sources are late, and those that are have
a greater value than Farnell admitted. It is true that, when the
Greek city-states dissolved into the cosmopolitan empires of
Macedon and Rome, there was a copious influx of oriental
cults, but these were to a large extent of the same ultimate
origin as the Greek. In Greece the prehistoric religion of
Anatolia and the JB gean, affected from the outset by Baby-
lonian and Egyptian influences, developed along distinctive
lines determined by the particularisation of the city-states; and
the result was that, when these states lost their independence,
their religious superstructures collapsed along with them,
leaving them once more exposed to the influence of the less
differentiated cults of Anatolia and the East. In this way the
ground was prepared for the mystical eschatology of the later
Orphics and Neoplatonists, which, elaborated though it was
with speculative novelties, obscurantist and sophisticated, was
essentially a revival of some of the most primitive elements in
ancient religion.
The weakness of Roscher’s position lies not in his sources
but in his treatment of moon-worship as a ‘thing-in-itself’
without reference to the structure of primitive society. The
lunar associations which we find clinging to nearly all the
146 Roscher SV.
VI THE MAKING OF A GODDESS 229
Greek goddesses and many of the gods are relics of the time
when it had been believed that the moon dominated the life
of woman. -
The deity most overtly associated with the moon was Hekate,
goddess of witchcraft. At the end of the month, when there
was no moon in the sky, the Greek housewife used to sweep
her floors and take the rubbish to a crossroads, where she
threw it down with averted eyes and returned without looking
back. Such deposits were known as ‘Hekate’ s suppers ’. 146 The
idea was that the human excreta swept up with the rest were
charged with magic and so dangerous. A comparative study of
the evidence makes this interpretation inescapable, and it is
confirmed from internal sources. An inscription from loulis
(Keos) records a law for the regulation of funerals. After
forbidding services for the deceased at the end of each
month — what Catholics to-day call the ‘month’s mind’ —
the law makes it an offence to place sweepings from the
house on the grave . 147 The prohibition of these practices
shows that they had been general. They sprang from the
belief that these monthly deposits helped the dead to be born
again.
On the sixteenth of the month, when the moon had just
passed the full, the women used to go to the crossroads and
offer to Hekate round cakes stuck with candles, which they
called shiners’ ( amphiphontes ). 148 The object was to preserve the
light of the moon. ‘Shiners’ were also offered to Artemis . 149
This is only one of many connections between the two deities.
iEschylus, whose authority is neither late nor dubious, speaks
of the moon as ‘the eye of Leto’s daughter', and elsewhere
he identifies Hekate and Artemis as a single goddess of
149 Harp. 6fuWwcc, A. C. 97 sch., Plu. M. 708-9, Ar. Pi. 594 sch., Poll. 5.
163, ThpKr. Char. 16. 7, Ath. 32.5a.
147 SIG. 12x8. 22.
146 Ath. 645a, Phot. <&u9i<jav. At Athens this rite became a commemoration
of the Battle of Salamis, which took place on the 16th of Mounychion (Plu.
M. 349Q, just as our All Hallows has become confused with the Guy
Fawkes plot: see further Jeanmaire 398-9.
149 In this ritual Hekate and Artemis were identified: Ar. Pi. 594 sch.
Cakes known from their shape as 'moons’ were offered to Selene, Hekate,
Artemis, and Apollo: Poll. 6. 76.
VI
230 STUDIES IN ANCIENT GREEK SOCIETY
childbirth .! 60 Hekate is in originsimply Artemis Hekate, sister
of Apollon Hekatos— the goddess who 'shoots from afar* the
shafts or pangs of childbirth.* 6 * Both goddesses are entitled
trioditis, referring to the crossways, the place where 'three ways
meet', and triprosopos , 'three-faced ’. 162 Since the crossroad
ritual was addressed to the moon,
the three ways were taken to repre-
sent the three lunar phases. ‘The
moon*, says Porphyry, ‘is Hekate,
who symbolises its changing
phases and the powers dependent
on them: that is why her influence
is manifested in three forms .’ 163
Porphyry was a Syrian Neoplat-
onist of the third century A.D.,
but more than a thousand years
earlier Hesiod had told how
Hekate was allotted 'a share in earth and sky and sea ’. 164 The
concept of a threefold Hekate was far older than Neoplatonism.
In Orphic literature the lunar phases are variously inter-
preted. r For the first three days,' according to one authority,
‘the moon is called Selene; on the sixth she becomes Artemis;
on the fifteenth Hekate .’ 166 ‘When she is above the earth*,
according to another, ‘she is Selene; when within it, Artemis;
when below it, Persephone .’ 166 These authorities are late, but
they were reproducing an ancient tradition. Epicharmos
identified the moon with Persephone on the ground that both
spend part of their time beneath the earth . 167 and Epicharmos
160 A. fir. 170, cf. E. Ph. 109-10; A .Su. 684-5. Famell dismissed this
testimony as a 'misconception 1 (CGS z. 460) — surely a strange approach
to the study of Greek religion. Elsewhere Aeschylus 'misconceives' Artemis
as a daughter of Demeter: A. fr. 333, cf. E. lo 1048.
161 Horn. H. 9. 6, ll. 7. 83, H. 3. 277, ll. 1. 14. On the Anatolian origin
of Hekate see Nilsson GF 397, PT 368. The name is very likely non-
Greek, the Greek interpretation of it being secondary.
162 Orph. HMag. 1. 1, Ath. 325a, Charicl. 1, Bergk PIG. 3. 682: fig. 14.
163 Porph. ap, Eus. PE. 3, n. 32.
164 Hes. Th. 427.
166 E Med. 396 sch.
iso Serv. ad Verg. A. 4. 511.
167 Epich. 54, cf. Serv. ad Verg. G. 1. 39, Cic. ND. 2. 27. 68.
VI THE MAKING OF A GODDESS 231
lived in the sixth century b.c. In fact the connection is already
implicit in the Homeric Hymn to Demeter. It is Hekate who
overhears Persephone’s cries when she is ravished — she does
not actually see the rape, only the Sun does that; it is Hekate
who meets the disconsolate Demeter and tells her what she
heard; and it is Hekate who embraces the resurrected girl and
becomes her faithful minister. 168 Who is Persephone) What
does her story, which has haunted poets ever since, really
mean)
9. The Rape oj Persephone
The cavern into which the women threw their pigs at the
Thesmophoria was called a migaron. This was the word regu-
larly used of caves sacred to Demeter and Persephone. 1 68 It was
also applied to a house or palace. That is its Homeric sense.
The chasm into which Hades disappeared with Persephone
was a megaron . 160
The earliest shrines were caves; the earliest dwellings were
caves. 161 Such was the Greek tradition, and archaeology confirms
it. 162 In palaeolithic Europe cave mouths and rock shelters were
used as habitations and caves as sanctuaries. In the neolithic age,
abandoned as dwellings, they remained in use as shrines,
tombs, and granaries. In Greece many of these cave sanctuaries
have yielded Minoan remains, notably the Cave of Amnisos
near Knossos, which is mentioned in the Odyssey . 163 The simplest
Minoan sepulchres are just caves and nothing more. There
were also artificial chambers dug out of the soil and enclosed
by monoliths, often with access through a door at the side or an
opening in the roof. 164 The megalithic monuments of western
188 Horn. H. 2. 24-6, 47-58, 438-40.
169 Luc. DMsr. 2. 1 sch., Plu, M. 378c, Paus. 1. 39. 5, 3. 25. 9, 8. 37.
8, 9. 8. 1, Clem. Pr. 2. 14, Eust. ad Oi. 1. 27, Phot, p&yapov. The
stone circles that served the same purpose (Paus. 2. 34. 10, 2. 36. 7) were
doubtless artificial plyapa.
166 Horn. H. 2. 379.
161 Porph. Ant. 20, Ps. Luc. Am. 34.
168 Childe DEC 4, 221, 231, 285 etc., Burkitt P. 90-1, 161.
163 Oi. 19. 188; Nilsson MMR 50-71. These caves, with their parapher-
nalia of figurines and other ritual objects, suggested the traditional des-
cription of the cave of the nymphs in Oi. 13. 102-I2,cf. 12. 317, Longus 1.4.
164 Childe DEC 50-1, 67.
VI
232 STUDIES IN ANCIENT GREEK SOCIETY
Europe are regarded by some archaeologists as reproductions
of natural caves. 188 Artificial tombs began as models of caves and
developed by following the pattern of domestic architecture. 188
The tomb was the house of the dead.
In Anatolia the Phrygians, and probably the Hittites before
them, used pits excavated in natural mounds and propped
with timber as refuges from the heat and cold. 187 Similar pits,
employed as granaries and entered by a ladder from the roof,
were common in Cappadocia, Armenia, Italy, Germany, Libya,
and Spain. In Latin they were called ‘wells’ (putci). Varro says
that cereals stored in this way, and firmly sealed, would keep for
many years — wheat for fifty, millet for more than a hundred. 1 88
The Roman mtrndus was a structure of this type. 189 At the
foundation of the city a pit was dug in the centre of the site
as a receptacle for the firstfruits. It was opened annually on
August 24 to receive the seed-corn from the harvest, and again on
November 8, when the seed was taken out for the sowing. The
unsealing ceremony was a solemn one. It was as though a door
were being opened to the spirits of the dead. As Jane Harrison
observed, ‘the same structure is treasury, storehouse, tomb:
ghosts and the seed-corn from the outset dwell together’. 170
Grain was stored at Eleusis. 171 Many Greek states used to
send their firstfruits to the Eleusinian Demeter. There they
were sealed in subterranean granaries till the autumn, when
they were taken out and sold. 172 For what purpose were they
sold? ‘Surely’, Cornford remarks, ‘not to be eaten, but to be
mixed with the grain of the sowing, like the sacra of the
Thesmophoria.’ 173 This was simply a ritual survival of the
ordinary procedure of storing the seed-corn.
The Laughterless Stone at Eleusis was so called because -
Demeter sat down there and wept. It corresponds to the Stone
188 Childe DEC 50—1, 209.
188 A. J. Evans PM 1. 72, Pendleburv 63, Xanthoudides 135.
18 7 Vitr. 2. 1. 5.
188 Varr. RR. 1. 57, X.An. 4. 5. 25, D.S. 14. 28, Tac. G. 16,' cf. D. 8.
45, Artem. 2. 24.
iGO Fowler MP, Harrison SID.
170 Harrison SID 143. The mundus was consecrated to Dis Pater and
Proserpina: Macr. Sat. 1. 16. 17-8.
171 SIG. 83. 11. 172 Harrison SID 145. 173 Cornford AEM 164—5.
VI
THE MAKING OF A GODDESS
233
of Invocation at Megara, where the goddess summoned her
daughter' from the dead (p. 1 31). In the Homeric Hymn the
Laughterless Stone is not mentioned. There Demeter is
described as sitting down beside the Maiden’s Well, 1 ’ 4 This
was not a natural spring (krint) but a phre'ar, a cistern or artificial
pit, the Latin puteus . Comford inferred that the Maiden’s
Well was properly a granary and the Laughterless Stone its
lid. 176 Another name for the Maiden’s Well was the Well of
Flowers, in allusion to the nosegays that Persephone was
gathering when she was carried off. 178 This then was the very
spot at which the rape took place — a subterranean granary, a
house of the dead, a threshold of the underworld.
The Eleusinian Mysteries were celebrated in September.
The sowing season, in the historical period,
was October, but in early times it had
begun a month before. 177 The harvest was
in June. The seed-corn was thus stored for
four months, a third of the year; and
this, in the myth, is the annual term
which Persephone has to spend in
the underworld.
For these reasons Comford
interpreted the rape of Perse-
phone as a symbol of the custom-
of storing the seed-corn from
harvest to sowing in underground
pits. 178 The sanctity that
attached to these granaries in
virtue of their immemorial as-
sociations with shrines and
sepulchres engendered the belief
that the grain so stored was
ferdlisedby contactwith the —■
dead, and the whole thing was nG< I5 ‘ in Haics: Attic vase
174 Horn. H. 2. 99. 176 Comford AEM 161.
176 Paus. 1. 39. 1. In the Sicilian version the rape was located at a spring
near Enna: D.S. 5. 4.
177 Plu. fir. xi. 23, Prod, ai Hes. Op. 389=Carm. Pop. 50.
778 Comford AEM 157-91.
VI
234 ' STUDIES IN ANCIENT GREEK_ SOCIETY
projected as a myth of the girl ravished by Hades, mourned by
her mother, and released on condition that for a third of each
year she returned to her subterranean lover. Persephone is the
spirit of the corn, which is buried and rises again.
This interpretation shows Cornford at his best — a bold and
clear-sighted materialist. But there is a small residue of details
which it does not cover. One of these he pointed out himself.
If Persephone was raped at harvest time, how was she gathering
spring flowers? He explained this as a later element ‘due to a
ritual enactment of the whole story in spring ’. 178 It is not hard
to find parallels. The Greeks of southern Italy had a festival
of Demeter and Persephone at which the women gathered
wild flowers for their hair . 180 A similar festival, the Erosan-
theia or Feast of Spring Flowers, was observed in the Pelopon-
nese . 181 But why should such a ceremony have been associated
with the spirit of the seed-corn?
Let us read the opening of the Homeric Hymn:
Persephone was playing with the nymphs and gathering flowers in a lush
meadow — roses, crocuses, violets, irises and hyacinths, and the narcissus
too, which. Earth had brought forth at the command of Zeus to please the
Lord of the Dead, who desired it as a snare for the handsome girl. . . . And
as she stretched out her arms in wonder to pluck this pretty plaything, the
ground opened and the Lord with his immortal horses leapt upon her. 182
There were other versions. At Megalopolis she was represented
as gathering flowers, but with Athena and Artemis . 188 At
Olympia her companions were the nymphs, but they were
playing ball . 184 Ball dances were common in Greek ritual. The
Spartan ball fight ( sphairomachfa ) was performed by boys at
initiation ; 186 and on the Acropolis at Athens a ball court was
reserved for the Arrhephoroi , 186 the girls who carried the box
down to the cave at the foot of the hill (p. 222).
We need not enquire too curiously into the botany of the
Homeric Hymn. The poet has evidently chosen his flowers
170 Cornford AEM 166. iso Str. 256.
181 Hsch. fipoa&vQeta, cf. Paus, 2. 35. 5. iss Horn. H. 2. 5-18.
I 88 Paus. 8. 31. 2, cf. Hyg. F. 146. isi p aU s. 5. 20. 3.
185 Eust. ad Od. 8. 376 =FHG. 2. 69, Paus. 3. 14. 6, cf. Caryst. 14=
FHG. 4. 359, Ath. I4d. I suspect that these ball-dancers had once been
girls: sec pp. 272-3.
• iso Plu. M. 839b.
VI THE MAKING ‘OF A GODDESS 235
with an eye to their intrinsic beauty. But the last of them
stands apart.- Like the hyacinth and crocus, 18 ’ the narcissus
was sacred to Demeter and Persephone, whose votaries plaited
it into garlands . 188 This mythical flower-gathering rests on
herbal magic. Every spring the women went into the meadows
to gather plants for their dyes, medicines, and spells. So far
from being late, this part of the myth is one of the earliest.
Herbal magic is older than agriculture.
As described in the hymn, the rape is a marriage by capture
— a patriarchal union, implying that the bride went to live
with her husband. But in that case where was the girl’s father,
ihd why did he not come to her rescue? Her father was said to
be Zeus , 189 but there is no hint of this in the hymn. The truth
is that originally Persephone was fatherless. So was Demeter.
It was her mother Rhea who gave her her name and conveyed to
her the promise to found. mysteries in her honour . 190 Similarly,
at Eleusis it is the queen who invites her into the palace and
entertains her . 191 The background of the myth is matriarchal.
By this I do not mean that the marriage itself is a late accre-
tion, only its form. The Eleusinian Mysteries included a
sacred marriage, enacted by the hierophant and high-priestess . 192
Details are lacking. In the Phrygian mysteries of Sabazios,
derived probably from those of the Hittite mother-goddess,
the priestess slipped a gold snake down through her vestments
to the ground — the ‘god through the bosom ’. 193 This gives us
the genuine matriarchal form of the ceremony, except that
originally, we must suppose, it was not enacted with a gold
18 ’ Hsch. SouArpiov, Paus. 2. 35. 5, S. OC. 683-5. The hyacinth was
believed to promote puberty: Plin. NH. 21. 170.
188 S. OC. 683-5.
189 Hes. Th. 912-3.
190 Horn. H. 2. 122, 459-69.
191 Horn. H. 2. 169-230. The kernel of the Mysteries was probably a
matriarchal palace cult (Deubner 88-91) like those discussed above pp.
124-5, 193-
192 Aster. Horn. io=Migne 40. 323, Ps. Orig. Plilos. 5. 1 = Cruice 170,
Clem. Pr. 2. 13. The women's rite in the cult of Demetcr at Sikyon was
enacted in a ‘bridal chamber’: Paus. 2. II. 3.
199 Clem. Pr. 2. 14. Some cults of Persephone seem to have included a
rite of this kind: Head 476.
VI
236 STUDIES IN ANCIENT GREEK SOCIETY
snake in a temple but in one of those prehistoric cave sanc-
tuaries in which real snakes abound.
Hekate overheard the girl’s cries but did not actually see the
rape. She was in her cave at the time — invisible. 194 She ap-
peared, torch in hand, nine days later. 196 During the interval
Demeter had been searching for her daughter and crying at the
crossroads (p. 131). These nine days, when Hekate was in-
visible and Demeter on her wanderings, are the period of the
dying moon, the last third of the month, the time for the
crossroad ritual. 186 And Persephone remained in the nether
world for a third of the year. 197 It seems that Epicharmos was
not far from the truth when he identified Persephone with the
moon.
In the neolithic age there was no solar calendar. This was a
much later invention, presupposing a well-organised priest-
hood and an official cycle of agrarian festivals. 198 Even in
historical Greece the clan cults, devoted to the ancestors,
retained their lunar basis (pp. 1 12, 125), and when we go back
to Hesiod we find the life of the peasantry regulated almost
exclusively by the moon. The annual storing of the seed-corn
had been grafted on a more primitive observance which was
not annual but monthly. 199
Why did Demeter mourn and search for Persephone? In
the circumstances created by the myth her behaviour is so
natural that the question may seem superfluous. But in these
matters the questions that most need to be asked are precisely
those which the story takes for granted. The answer is to be
found, I believe, in what has been recorded of the Mohawks
of North America:
When a young woman finds herself come to a state of maturity, she retires
to'conceal herself with as much care as a criminal would take to keep out of
the reach of justice; and when her mother or any other female relative
194 Horn H. 2. 25. "6 Horn. H. 2. 51-2.
196 i n western forms of the Thesmophoria the period of abstinence im-
posed on the women was nine days: Ov. Met. 10. 431-5, cf. D.S. 5. 4,
Orph. fir. 47.
197 Horn. H. 2. 398-400. 198 Nilsson PT 173, 231-2.
199 The change was assisted by die fact that the early Greeks recognised
only three seasons (Nilsson PT 71-2); or perhaps we should say that the
three seasons were modelled on the three enneads.
VI
THE MAKING OF A GODDESS
237
notices her absence, she will inform her female neighbours, and all will begin
to search for the missing one. They are sometimes three or four days without
finding her, all of which she passes in abstinence, and I really believe she
would rather die than show herself before they find out. 200
Lastly, before leaving her house of death Persephone was
induced to eat a pomegranate seed . 201 We have learnt what this
means. Thereby she condemned herself to a periodical return.
In the story as we have it this is an annual period of four
months, but originally it was the menstrual period of the
dying moon.
Who then is Persephone? Is she a moon-goddess, as Roscher
maintained? Is she a corn-maiden, as Cornford proved? Is she
a queen of the dead, as she was to her ancient worshippers?
She is all these — 'goddess and maiden and queen' — but she
is also an ordinary young woman, embodying the actual ex-
perience of girlhood from the daughters of the paleolithic
cave-dwellers, brutish in their looks and filthy in their habits,
to the smartly-dressed young ladies that made such a fine show
at the Athenian carnivals.
10. The Female Figurine
The oldest extant piece of statuary was discovered in an
upper palaeolithic loess deposit in Lower Austria. It is
carved in soft oolite limestone, eleven centimetres high, and
represents a nude woman with the arms
folded across the breasts. It is known as
the Venus of Willendorf. To those who
have admired the Venus of Milo in the
Louvre the title may seem inapposite. This
palaeolithic Venus is fat, thick-hipped, heavy-
breasted. She was painted with red ochre . 202
These female figurines have turned up
in hundreds among the neolithic and
chalcolithic deposits of Central Europe, the Mediterranean
region, and the Near East. They are usually made of baked
clay (terracotta) but sometimes carved in stone. Male figures
200 D. Cameron quoted by Briffault 2. 3^9- 201 Horn. H. 2. 37 * — 4 -
20= Macalistcr 1. 447, Burkitt P 222: fig. 16.
FIG. 16.
Venus of Willendorf:
paleolithic figurine
VI
238 STUDIES IN ANCIENT GREEK SOCIETY
occur, though they are less common than females, and also
models of animals.
In Phase I of the Danubian culture a small quantity of
female figurines have been found. In Phase II they are abundant.
In Phase III they disappear. Phase III is marked by the develop-
ment of stock-breeding and warfare (p. 34). 203 If these changes
were accompanied by a decline in the status of women, as they
normally are, that would account for the disappearance of the
female figurines. 204
The Gumelnita culture of Rumania is rich in ritual remains.
Phase I includes a large number of well-modelled clay figurines,
all female. They continue into Phase II, but males too are now
represented, together with clay phalli. Above the Gumelnita
deposits lies a later culture, distinguished by flint arrow-heads
and battle-axes. In this there are no female figurines. 206
A similar sequence has been established in neolithic
Thessaly:
In general all the earlier figurines are well made of refined
clay, usually polished and in some cases painted in the
red-on-white style or something akin to it. The majority of
the human figurines are female; a few are male. . . . Most
of the early female types are very corpulent, with anatomical
details greatly exaggerated. . . . They are represented standing
or sitting, sometimes with one foot under the body. The
arms are extended beside the hips, folded across the body,
or support the breasts. 200
In the Dimini culture (Thessalian II) figurines are
still found, in stone, as well as clay, and two new
types appear — a seated woman with a baby in her
and a seated man, ithyphallic, with his hands on his
knees. There are also models of cattle. But the Dimini figurines
are less plentiful than those of the preceding period, and
inferior in execution. This deterioration continues in Thessalian
III, and after that they disappear. 207
The Minoan figurines have been described by Evans. Frag-
ments of male figures have been found, but die great majority
are female. The commonest type, which has its nearest parallels
in Anatolia, is short, stumpy, and steatopygous. Evans adds:
sis Childe DEC 99-108. 204 io 8. 206 2J. 126-9.
206 Hansen 43-4. 2 o7 2 fc. 68-71, 91: fig. 17.
HG. 17.
Thessalian
figurine:
terracotta from
Sesklo
arms,
VI
THE MAKING OF A GODDESS
This evidence points to the existence already in the neolithic age, both on
the iEgean and Anatolian side, of related families of squatting or seated
female figures formed of clay and of obese or steatopygous proportions. The
appearance of one of the stone offshoots of this family as far east as the
middle Euphrates is a phenomenon of the greatest interest in connection
with the diffusion of a parallel group
of female figures through a wide
Semitic region and even to the seats ■ '
of the Anau culture in southern Tur-
kestan. . . . Among the earliest known ’
examples of this oriental class are the
clay figurines, identified with the
Babylonian mother-goddess, found at
Nippur and dated about 2700 B.c. 208
Thus, while the figurines of S.E.
Europe and S.W. Asia de-
veloped to some extent under
Babylonian influence, it is clear
that the image of the Babylonian
mother-goddess herself had
evolved from the same origin.
There is consequently no reason
for postulating a Babylonian
origin for the actual cult, which
is characteristic of the whole
domain.
‘In Crete itself’, Evans goes
on, ‘it is impossible to dissociate
these primitive images from
those that appear in the shrines
and sanctuaries of the great
mother-goddess . * 2 0 0 In Greece fig. 18. Minoan figurine: terracotta
too throughout the neolithic and f rom ^ n0S50S
chalcolithic periods we find these female figurines in abundance,
and, since the Greek goddesses of historical times have admitted
affinities with the Minoan, we are obliged to infer that they go
back to the same neolithic prototype.
During his excavations at Mycense — the first to be under-
taken on this site — Schliemann discovered, mostly in tombs,
a large quantity of female figurines. They are so crudely
288 A. J. Evans PM 1. 45, 51. 209 *• 5 -•
VI
240 STUDIES IN ANCIENT GREEK SOCIETY
modelled that he mistook them for cow’s heads, which he in-
terpreted as symbols of the ‘cow-faced’ Hera. 810 This was an
error, but there is no reason to doubt that they represented
Hera. A temple of the same goddess at the mouth of the Sele in
southern Italy has yielded over 200 terracotta statuettes of a
female figure. 811 They are much later in date and quite different
in style from the Mycenean, but they must have been inspired
by similar motives, and there can be no question that they are
effigies of the goddess. One of their commonest attributes is a
basket of pomegranates held in the hand. Archaic figurines
have also been found at Heraia near Corinth and in the temple
of Artemis at Sparta. 812
It is agreed that these objects were intended somehow to
promote fertility. So much indeed is in some specimens
obvious. But there the problem has been left. In attempting
to solve it, several considerations must -be kept in mind. In
the first place, cults involving the use of human effigies are
not confined to this region of the world nor to the past.
Secondly, since the remains present a continuous series from
Late Palaeolithic to the Iron Age, we must be prepared to
find that they served different purposes at different times.
Between the first and last of them lies almost the whole
history of magic. Thirdly, the circumstances of discovery
demand attention. Most were found in tombs; many of the
later examples must have been votive offerings, some being
perforated for suspension. In some cases too the postures -and
gestures are obviously intended to be significant. And lasdy,
the sequence revealed in the stratified cultures of the Danube,
Gumelnita, and Thessaly suggests that they should be- studied
against the background of the primitive agricultural matri-
archate.
A number of female figurines, modelled in clay, with ex-
aggerated sex-marks, have been recovered from neolithic
deposits in Japan. 813 Female statuettes, carved in wood and
highly stylised, have been found in the PhilippineTsIands and
210 Nilsson MMR. 260-2. an Zanotti-Bianco 244.
212 p a yne P 197-227, Dawkins 145-62.
213 Matsumoto 58, Adam 111-2. Also on upper palaeolithic sites in E.
Siberia: De Pradenne 191.
VI THE MAKING OF A GODDESS 24I
the south-easterly Carolines. 21 * In parts of Africa an important
part in women’s ritual is played by wooden dolls. It is here,
where the ritual context is still alive, that we must begin our
search.
Savage children, like our own, play with dolls. Among
savages, as among ourselves, these playthings are mainly for
girls. Playing with dolls, as we can all see, is a rehearsal for
motherhood. But it is a much more serious business among
savages than it is in a modern nursery. Unfortunately we know
very little about it. The male savage does not pay much atten-
tion to -it — it would be very improper if he did — and the male
anthropologist even less. Moreover, even if asked, native
women are not likely to reveal to white-skinned university
professors the secrets they guard from their own husbands;
and the social status of our own women does not encourage
them to take up anthropology. The result is that the woman's
half of primitive life, which for the study of origins is the
more important half, is very poorly documented. Special credit
is therefore due to Miss Earthy for her full and frank account
of the initiation of girls among the South African Valenge . 215
Early in spring the local chief issues a proclamation sum-
moning all the girls who have reached puberty during the year
to be initiated. The ceremonies that follow last a month. They
are superintended by a woman called the nyambutsi, who has
inherited the office from her mother, and with it the initiatory
symbols, which have been handed down in a special basket
from mother to daughter for generations. They consist of
a drum, a'horn, models of the genitalia of both sexes, and male
and female wooden dolls, all painted with red ochre. On the
first day, when the candidates have assembled, a band of
initiated women, led by the nyambutsi, perform a nude dance
to the beating of the drum, which is a symbol of the womb.
Meanwhile, as they watch, the novices are sobbing bitterly,
overcome with terror. In the evening, when the dance is over,
each girl submits in turn to an operation in which die hymeneal
membrane is pierced with the sacred horn. On the succeeding
si* Boas 69, Adam 125.
21s Earthy Ul-24. On the need for more women in social anthropology
sec Hambly 2S4, Ehrcnfcls 63.
Q
242 STUDIES IN ANCIENT GREEK SOCIETY' VI
days they receive methodical instruction in the facts of sexual
life. It is for this purpose that the dolls are produced from the
basket. Together with the genital images they serve as working
models of the sexual act. They are treated with great venera-
tion, because they are supposed to be vehicles for the activity
of ancestral spirits. During this time the novices are taught
a secret language, and are encouraged to steal from one another,
which they may do with impunity. On the last morning of the
month the dance of the first day is repeated, all the performers
being now covered in red ochre. But this time the girls do not •
weep. They beg the nyambutsi to open her basket for the last
time, and when she complies they dance round the dolls in
delight, clapping their hands and singing: ,
Babies elect, babies elect,
Babies, we greet you because you are beautiful!
All is now over. The nyambutsi packs up her treasures, the girls >
_ go home and take off their ornaments, which their mothers
• stow lovingly away in some secret corner of the hut.
Magic is a mimesis — a rehearsal or make-believe; and games
are an offshoot of magic. In children’s games the make-
believe has become an end in itself, but in magic it is directed
consciously to a practical object. When a peasant makes a wax •
image of an enemy and sticks pins into it or melts it over the
fire, he is engaged in primitive magic. The Valenge dolls
serve the same purpose. Objectively, they are instruments for
demonstrating the actual practice, but subjectively the de-
monstration assumes the character of a magical rehearsal.
The magical element cannot be separated from the actual
technique; it is simply its subjective aspect. At the end of the
demonstration the dolls are no longer man and woman; they
have become infants, fulfilling the promise of die demonstra-
tion and conveying the further promise that in due course the
whole process will repeat itself in real life . 216
Returning to the figurines, we see from the early preponderance
216 Himmelheber, quoted by Adam 97, reports from the Ivory Coast
that a woman may sometimes be seen carrying a doll on her back ‘to bring
home to her body that she wants a child like that.' Wax images of. the
‘Mother’ are still dedicated in the Tyrol as a cure for sterility: McKenzie
298.
VI
THE MAKING OF A GODDESS
243
of females that they cannot have been designed originally
for the same object as the Valenge dolls. They go back to the
time when the connection between copulation and conception
was unknown, as it still is among the lowest savages. 21 ’ The
magic for which they were made was directed in the first place
to menstruation and parturition, and extended later to initia-
tion, marriage, disease, and death — to every crisis that de-
manded the infusion of reproductive energy, the renewal of life.
The presence of phalli along with male figures
in the later deposits suggests that this stage corre-
sponds to the Valenge ritual. The figures are sail
made by and for women, but as puppets for
demonstrating the sexual act. The third stage,
marking the development of anthropomorphic
divinities, brings us within range of the Greek
data. Being associated with a deity who is
imagined as a woman, the effigies become
confused and identified with her. The figurine
becomes a cult statue.
- The earliest Greek figurines are the Cycladic.
They occur mostly in tombs, and the charac-
teristic type is a nude woman with the arms
crossed beneath the breast . 218 Specimens of this
■ type were imported into Crete, where they have
been found in Early Minoan tombs. The Middle
and Late Minoan figurines fall into three classes
• — those found in tombs, votive offerings from
sanctuaries, and cult idols. In’ the females the
hands are almost invariably held beneath or
before the breasts, sometimes with one of them
raised. In this they anticipate the so-called
‘dancing girls’ — bronze statuettes in flounced
skirts with one hand against the forehead and
the other on the waist — what Nilsson has called
FIG. 19. Cyclaite
figurine: marble
217 Their ignorance is nor surprising when we realise that 'there is no
such thing as a virgin among die native tribes of Australia’ (Spencer
NTNTA 25): sec p. 2S7 n. 182. Even where the process is understood, the
sexual act is often regarded simply as the medium through which the
woman is impregnated by animal or vegetable spirits: Karstcn 427-9.
2 » 8 Nilsson MMR 251: fig. 19.
VI
244 STUDIES IN ANCIENT GREEK SOCIETY.
the 'gesture of benediction'. The same attitude is found in
some of the male specimens; in others the hands are placed
on either side of the chest . 219
The Mycenean specimens, also found in tombs and shrines,
are divided by Nilsson into three types. In the first the head
has a cap, with the hair flowing down
the back; the arms are mere projec-
tions, like horns — hence Schliemann’s
mistake. The second is capless
and armless. In the third the
arms are resting against the breast
and sometimes crossed . 220
These postures are undoubtedly
I symbolical. At Aigion there was
a cult statue of Eileithyia in
which one hand was extended
straight forward while the other held
up a torch . 221 The same goddess
appears in the same posture in a vase-
painting of the birth of Athena , 222 and
in poetry Artemis is described as holding
both hands over a woman in labour . 223
This gesture, as Farnell observed, was
supposed to assist childbirth. The reverse
sign, for retarding delivery, was to lock
the hands by intertwining the fingers . 224
no. 20. ‘The gesture of Having established this point, we may
benediction^ bronze from i n f er ^ ver y comm0 n sitting or
nossos squatting attitude represents the actual
moment of delivery. We know that among savages the woman
squats or kneels, supported by the midwives, and we have
220 Ih. 260-2.
222 Farnell CGS 2. 614.
210 Nilsson MMR 252-6: fig. 20.
221 Paus. 7. 23. 6.
223 AP. 6. 271.
—4 Ov. Met. 9. 292-300, Plin. NH. 28. 59, cf. Ant. Lib. 29, II. 19. 1x9,
Paus. 9. 11. 3. Women entered the temple of Juno at Rome with all knots
untied: Or. F. 3. 257-8. It is still customary to unfasten knots and locks
in a house where a woman is in childbed: Frazer GB— TPS 294-8. Among
the Bathonga ‘no knot must enter a grave' Qunod LSAT 1. 140)— in order
that the dead may be born again.
VI THE MAKING OF A GODDESS 245
some archaic Greek statuettes of kneeling women, Eilcithyiai
or Genetyllides, goddesses of childbirth. 826
The Jukuns of Nigeria have a rain dance, performed by the
king’s daughters. It is very simple. The dancer moves her
hands alternately and repeatedly from head to hip and hip to
head — the ‘gesture of benediction’. This parallel is all the
more interesting because it is believed that the Jukun priest-
kings have historical connections with ancient Egypt. 220 It
suggests that the varied attitudes struck by the figurines may
have been passing moments in a dance — ‘the sacred chant of
Eileithyia’ 227 — and at the same time it confirms the view that
such gestures were considered to be as efficacious for the growth
of crops as they were for the birth of man. These puppets
could render assistance in all the vicissitudes of life, including
the last. That is why so many of them have been found in tombs.
There remain the specimens from sanctuaries, which
Nilsson classes as votive offerings. This term had better be
avoided for the moment, because it prejudges the issue. No
doubt many of them, at least in die later deposits, were
votives, but not all offerings are votive, and their neolithic
antecedents must have belonged to pre-deistic cults in which
the very idea of an offering was unknown. 228
The Greek cave sanctuaries have been prolific in ritual re-
mains, and the nature of the ritual is apparent from die associ-
ated traditions — the cave at Amnisos in which Hera gave birdi
to Eileithyia, 228 the caves in which Rhea gave birth to Zeus, 280
and the ubiquitous caves of the nymphs. 23 1 Many contain pools
or springs. Sacred springs were as plentiful as sacred caves, and
they tell the same story — the springs where Rhea was purified
after her delivery, 232 the Maiden Springs, the springs in which
girls bathed before fesdvals (p. 224), and die springs consecrated
225 Earthy 69, 71, Roscoe B (1911) 5 *i SB 242, BTUP 24, cf. Hutton
233; Famell CGS 2. 613-4, cf. Horn. H. 3. 116-8.
220 Meek 283, 191, 196, 202, 207.
227 Call. Del. 257, cf. Ov. Met. 9. 300-1.
228 Many sepulchral deposits usually classed as votives are more probably
charms: Karstcn 244-5, 251-2.
228 Oi. 19. 188, Str. 476, Paus. 1. 18. 5.
aao paus. S. 36. 3, Call. /ov. 10, cf. A. R. 4. 1130-6.
233 Roschcr LGRM 3. 509-11. 529-54- 232 P«»- - s - -•
VI
246 STUDIES IN ANCIENT GREEK SOCIETY
to the nymphs . 288 From time out of mind these places
had been the scene of women’s mysteries in connection with
initiation, menstruation, marriage, childbirth, and hence it is
not surprising that they have yielded so many figurines.
As cult properties, die figurines were naturally stored in the
sanctuaries, where there was no risk of their being damaged
or causing damage through the magic they contained. When
the idea of a goddess took shape, they came to be regarded as
hers and as deriving their potency from her. Similarly, the
women who absorbed their magic by manipulating them be-
lieved themselves to be filled with the divinity and so identified
with her . 234 By this time the figurines had become something
more than puppets. They were regarded indifferendy as repre-
sentations of the goddess herself or of her worshippers. This
enables us to understand how they came to be used as offerings.
A votive offering properly so called is made in fulfilment of a
vow. You are in a fix, so you promise God that if he will get
you out of it you will give him this or that. The payment is
frequently made in advance. The votary may flatter himself
that this is a mark of faith on his part, but really it belongs
to a more primitive stage in the evolution of the custom. In
Greece, when the catde were diseased, the farmers used to
make models of oxen and dedicate them in the temples : 286
These are my oxen; they helped me raise my crops. They are only made of
dough, but take them kindly, Demeter, and vouchsafe in return that my
real oxen may live and fill my fields with sheaves. 233
Why does the deity get such a poor bargain? The offering of a
replica in return for die original cannot be explained in terms of
propitiation. It belongs to mimetic magic. My enemy prospers,
233 Roscher LGRM 3. 509-11.
234 The same process can be followed in the evolution of portrait statues
out of images carved from the sacred tree. The image of Artemis Lygodesma
was made of her own withy: Paus. 3. 16. 11. The old statue of the Argive
Hera was of pear-wood: Paus. 2. 17. 5, cf. Plu. M. 303a. The image of
Asklepios Agnitas was of agnos: Paus. 3. 14. 7. The initial stage is seen in
the Corinthian cult of Dionysus, who before a statue was made for him had
been worshipped simply as a tree: Paus. 2 . 2. 6 . 235 Farnell 2. 579.
230 AP. 6. 40, cf. 55* On the same principle models of parts of the body
were dedicated by patients in thanks for their recovery: Rouse 21 1. Many such
objects have been found in Minoan cave sanctuaries: Nilsson MMR 63, 69.
VI
THE MAKING OF A GODDESS
247
so I make a wax image of him and burn it. My cattle are ailing,
so I make models of healthy cattle. This is the point of the
whole business. 33 ’ The dedication of the image is, so to speak,
an afterthought, prompted by the consideration that being
charged with magic it needs to be put away in a safe place.
So widi the figurines. As representations of the worshipper
they were dedicated, by her in order to place her under the
goddess’s protection. This was done both in times of actual
danger, sickness or childbirth, and in times of imaginary
danger, such as initiation, marriage, or bereavement. As repre-
sentations of the goddess, they were an acceptable gift in return
for her favours. The interpretation was immaterial. The im-
portant thing was the rite itself, the dedication, which had
taken tire place of the magical act.
In the market-place at Troizen Pausanias saw a row of statues
of women and children erected by Athenian women who had been
evacuated there during the Persian invasion . 338 In this case the
motive was retrospective — gratitude for survival. But wc also
hear of statues erected to persons absent or missing with the
object of ensuring their safe return . 338 Here the idea of magic
is still active. In the same way the original purpose behind the
practice of erecting statues to the dead was not to perpetuate their
memory but to perpetuate their existence in the spirit world.
Throughout Greece it was the custom to dedicate statues of
victors in athletic contests and priests and priestesses at the ex-
piry of their office. How the athlete came to be regarded as
divine will be considered when we investigate the Olympic
Games. The priests and priestesses derived their sanctity
directly from their function. The Athenian Arrhephoroi, for
example, held office for a year, and on their discharge statues
were erected in their honour . 2 * 0 These were portraits of the
237 The oldest example known to me is in the Babylonian Epic cj Crestien
I. 61-5, where Ea sends Apsu to sleep by making a model over which he
recites a hypnotic incantation: S. Smith 37.
=38 p a us. 2. 31. 7. 230 Bcnvcnistc SMK.
2 «IG. 2. 1378-85, 1390-2, 3. 887, 916-S, cf. Paus. 2. 17. 3, 2, 35.
8, 7. 25. 7, Hdt. 2. 143. All the statues on the Athenian acropolis were
dedicator}-: Paus. 5. 21. 1. The sanctity of these Arrhephoroi appears a!so in
the rule that any gold ornaments they might wear became sacred to the
goddess: Harp, ipprjjvptlv.
248 STUDIES IN ANCIENT GREEK SOCIETY VI
girls themselves but as representatives of the goddess whose
nature they had assimilated through contact with her sacra.
Greek sculpture never entirely divested itself of these magical
associations.
One of the earliest examples of archaic sculpture is a marble
statue from Delos. It represents a standing woman, with flowing
hair, arms down the sides, and the body draped in a long
chitSn. Beneath it is an inscription:
Nikandra dedicated me to the goddess who shoots her arrows from afar,
Nikandra, the peerless daughter of Deinodikes of Naxos, sister of Deinom-
eneus, and now die wife of Phraxos. 241
As the last words show, it was dedicated at marriage. Whom
does it represent? Is it Artemis or Nikandra? Perhaps Nikandra
herself could not have answered that.
And so at last, in Greece as elsewhere, we reach the final
chapter in the long history of the Venus of Willendorf.
Strolling one sunny day along the shady banks of the Uissos,
Sokrates and Phaidros came to a shrine of the nymphs, where
they saw a number of votive offerings — images and dolls. The
Greek for a doll is k 6 re, a ‘girl’. The sight was evidently a
familiar one, so they. merely noted it and passed on . 242 Who
had left the dolls there, and why? The answer is given in a
dedicatory epigram:
To thee, Artemis of the Marshes, maid to maid, as is meet, Timarete
presents, as bride-to-be, her drum, ball, and headband, her dolls and their
dresses. O Daughter of Leto, stretch thine arm over her and bless her and
keep her pure and safe from harml 242
Timarete is going to be married, so it is time to put away
childish things. But though only toys they cannot be just
thrown away — they must be returned to the goddess to whom
they have always properly belonged, because there still clings
to them a faint aura from the time when Timarete’s remote
ancestresses had handled the same symbols, vibrant with the
power of renewing life, in the damp moonlit darkness of a
paleolithic cavern.
24 1 GDI. 542 3 . In an epigram attributed to Sappho a statue dedicated in simi-
lar circumstances is described expressly as a portrait of the donor: AP. 6. z6 9.
242 pi, pldr. 229-30.
243 jip. 6. 280, cf. 189, 309, SIG. 1034, K. M. Eldcrkin JDA.
VII
SOME MATRIARCHAL DEITIES OF THE 4EGEAN
I. Demetcr
While the figurines were degenerating into dolls, the matri-
archal goddesses who had taken them over from primitive
magic were adapting themselves to their new patriarchal en-
vironment. The beginnings of this process can be detected in
the Late Minoan period.
The standard Minoan burial practice was collective inter-
ment in natural caves or tholos tombs. 1 At Mochlos and in the
Mesara the graves are grouped in cemeteries, implying die
congregation of several kindreds in a single village settlement. 2
Collective burial was also general in the Cyclades, Atdca, and
the Peloponnese. 3 In the Peloponnese it was exceptionally per-
sistent. One of the sepulchres in the Grave Circle at Mycenae
was in continuous use for two centuries (1450-IZ50 B.c.) and a
family likeness has been recognised in the skeletons. 4 Recent
excavations at Malthi (Messenia) have brought to light a Late
Helladic village comprising over 300 rooms variously grouped
in closely packed houses. It was fortified, and outside the wall,
near the main gate, was a large cemetery enclosed by mono-
liths like the Grave Circle at Mycenae 6
On die other hand, individual interment in jars, stone cists,
or clay coffins had already begun in Crete and die Cyclades as
far back as the Early Minoan period, and became increasingly
common.® The jar burials arc perhaps a special case, many of
them being designed for infants. It is a widespread custom to
bury infants in jars cither in the house or just outside ic with
the object of reimpregnating the modicr with the spirit of die
1 Childe DEC 22-3, A. J. Evans PM I. 70-2, Hall CGBA 44.
= Childe DEC 23, Pcndlebuiy 63-5. * Childe DEC 50-1, 67.
4 xb. 76, cf. 209. 8 Valmin SME.
c Childe DEC 24. 50-1, A. J. Evans PM 1. 149-50 •
250 STUDIES IN ANCIENT GREEK SOCIETY VII
dead child . 7 But in general individual interment must be taken
as a sign of the disintegration of the clan. It shows that the
Minoan and Cycladic cultures were already moving along the
road fro m tribe to state — a process which was subsequently
repeated, in different con-
ditions, all over Greece..
Of the Minoan tribal
system we know at present
nothing, but totemic sur-
vivals abound. At Praisos
there was a taboo on sow's
flesh, supported by a tradi-
tion that the infant Zeus
had been suckled by a sow . 8
The people of this district
wereEteokretes or True Cretans®
— that is, of Minoan stock; and,
since the name of Zeus is Indo-
European, we may infer that it
had been attached by Greek-speaking invaders to an indigenous
totemic cult. In other parts of the island Zeus was associated"
with the goat . 10 On the slopes of Mount Aigaion, the Goat
Mountain, which was consecrated to him, there is a natural
grotto, the Cave of Psychro, in which a Minoan vase has been
HG. 21. Trojan fare urn
7 Frodin 437. Among the Valenge 'all infants who are bom dead ot
die under the age of a few months are given a pot burial' (Earthy 153)
and 'a water-pot is one of the symbols of the womb' (66), In S. America
‘the clay jar in which Indians bury their dead may be taken to represent
the womb' (Karsten 34-5); 'the clay vessel is a woman, just as the earth
itself from which the day is obtained is regarded as a woman’ (246-7, cf.
251-2). Neolithic pots marked with a female head and breasts have been
found in Cyprus (Lang 187) and a woman's head is a characteristic design
on the so-called face urns of Anatolia (Childe DEC 41: fig. 21). All this
has a bearing on the myth of Pandora, which I hope to discuss in a later
volume. On the ritual of pot-making see Karsten 240-1, Briffault 1. 466-7.
8 Ath. 376a.
8 Staph. I2=JFHG. 4. 507.
10 He was reared on Mount Ida by die nymph Amaltheia, who fed him
on goats milk (Erat. Cat. 13. Hyg. Ast. 2. 13) out' of a cornucopia (Hsch.
■AfJc&tefos idpas). The livestock of Minoan Crete was mainly pigs and goats
(Childe DEC 21). 1 r &
VII
SOME MATRIARCHAL DEITIES
251
FIG. 22 . Child suckled by goat:
Minoan seal
found decorated with goats and double axes. 11 It would seem
that the goat rendered the same service as the sow, because
we possess a Minoan seal with the
design of a goat giving suck to an
infant. 13
The characteristic symbols of
Minoan religion are the pillar,
double axe, and bull's horns, and
die most important animals are the
serpent, dove, and bull. All these
come from Anatolia. The pillar is
a stylised form of the sacred tree. 13 The axe must have owed
its sanctity in the first place to its use for hewing timber, which
in primitive society is women's work. 1 * Being used for felling
1 trees, it was associated widi die
' lightning, and so became a rain
charm. Later still it became a
batde-axe and a sacrificial axe, and
in the last capacity its potency was
further enhanced by contact with
die blood of the sacred bull. 16 Of
the animals, die serpent and dove
have already been discussed (pp. 1 14
-20, 21 3). The bull, embodying die
reproductive energy of the male,
was the deified leader of the herd.
Catdc worship, which can be studied
fig. 23. Minoan double axe:
intaglio from Knossos
among modern pastoral communities, is attested for ncolidiic
Europe by bull figurines discovered among actual bones of
catdc. 1 ®
The Minoan modicr-goddcss was served by priestesses,
assisted by male attendants. On a Minoan signet three
priestesses arc dancing in a meadow. Their breasts arc
m Glotz CE 252-3, Nilsson MMR 56-S.
12 A. J. Evans KE SSt %. 22.
is Id. MTPC. , , , ,
14 Mason 135. The Minoan doable axe is never found in the hands cf a
male deitv: Pendlebuty 274- A characteristic motive in the neolithic culture
of the Tarn and Garonr.c is a svotnan carrying a double are: Child; DEC :as.
is Glotz CE 251-2. ie Child: DEC 1 37. Rctcoe T. «-■*.
STUDIES IN ANCIENT GREEK SOCIETY
FIG. 24. Mttioan bull fight: intaglio firom Knossos
open and they wear flounced skirts. Lilies bloom beneath
their feet. Above, in the distance, the goddess hovers in
mid-air; belqw, as though ascending to meet her, a snake
r r »* r
fig. 25. Mycenean cult scene: gold ring from Mycente
rises from the ground . 17 On a gold ring from Mycenas a
priestess bends over an altar in an attitude of lamentation,
while another is dancing, with her elbows bent and her hands
17 Glotz CE 249: fig. 71. The fruit trees on one of these gems have been
identified as pomegranates: Bossert 327.
VII SOME MATRIARCHAL DEITIES 253
just descending to her hips. To the right is a tree, which a male
acolyte is bending down for the priestesses to pluck die fruit . 18
Ac Gournia the goddess is represented in a cult idol as a woman
in labour. 1 ® This is the aspect in which she was worshipped at
the Cave of Eilcithyia at Amnisos. Wc may take it diat the
FIG. 26. Dance at a sacred tree: gold ring from Mycenc
Greek Eilcidiyia, whose name is not Indo-European, was
descended direedy from the Minoan goddess of childbirth.
Some scholars have gone further and suggested an etymological
connection between Eilcithyia and Elcusis . 00
In confirmation of the Minoan ancestry of Dcmctcr, which
FIG. 27. Ascension 0 f Dmcter: terracotta relief
has been discussed in Chapter IV, attention may be drawn to
some of her cult monuments. In the Louvre dicrc is a terra-
cotta relief of Dcmctcr rising from the ground with corn-
stalks in her hands and a snake gliding up cidicr arm. Its dare
18 Glont CE 23S: fie. 26. Hawes 11 pi. 10. Nilsson MMR4 $o-i.
254 STUDIES IN ANCIENT GREEK SOCIETY VII
is late, but the type is probably old . 21 In some early statuettes
found at Eleusis she wears a tall cylindrical hat like that of die
modern Gre.ek pappds , 22 These two types have nothing in
common with one another, but both recall some Middle and
Late Minoan statuettes, in which the goddess or her
priestess is represented in a tall hat, flounced skirt,
and a bodice open at the breast; the hands are
extended, and in one or both there is
a snake, while others are coiling round
the arms, shoulders, and head . 23 And
this Minoan type invites com-
parison with the traditional con-
cept of the Erinyes as women
with snakes in their hair and hands . 24
In Arcadia the goddess was actually
worshipped as Demeter Erinys . 28 It
would seem that she originated as a
particularised form of the 'Erinyes,
to whom she stood in the same sort
of relation as Artemis to the nymphs.
Persephone was known at Eleusis
as Kore, the Maid; at Andania
as Hagne, the Pure Maid . 28 -
FIG. 28. Pappas type
of Demeter: terracotta
from Eleusis
Ariadne, the legendary princess Mtoomnah
of Knossos, is Cretan Doric for priestess:
eridgne , the Very Pure Maid, 2 ? statuette
and her stoty — after being carried off by Theseus to Naxos
she was ravished by Dionysus, with whom she disappeared
21 Roscher LGRM 2. 1359, cf. A. J. Evans PM 3. 458: fig. 27.
22 Famell CGS 3. 215: fig, 28.
22 Hall CGBA 127-8, Pendlebury pi. 28. 2. These Minoan statuettes
may be compared with a Syrian figurine described by Pritchard 36: "The
nude female figure appears to be holding a serpent in die left hand; another
serpent is shown draped about the neck with its head pointing to the genital
region.'
24 A. C. 1046-8; Roscher LGRM 1. 1331-4.
22 Paus. 8. 25. 4.
22 SIG. 736. 34.
2 ? Hsch. ASvdv. The form ’ApiAyvri occurs on vases: Roscher LGRM
1 539. She was also identified with Aphrodite: PIu. Ties. 20.
VII
SOME MATRIARCHAL DEITIES
-55
into the mountains 88 — is the rape of
Persephone in another form. A third figure
of the same type is Britomartis, whose
name is one of the few Minoan words
that we can understand: it means the
'sweet maid ’. 28 Pursued by Minos for
nine months, she eluded him by plunging
into the sea . 30
The Minoan mother-goddess had a male
partner, her son or her consort, or both.
The god docs not appear in the neolithic
age at all. In the Middle and Late Minoan
periods his status rises, but he remains
subordinate to the last . 31 He stands for
the patriarchal principle emerging within
the matriarchatc. Ac Knossos he seems to
have been identified with the bull. Hence
the myth of the Minotaur, loved by
Queen Pasiphac, which is perhaps founded
on a sacred marriage, the male part
Pin. Jo. ilnitnatu:
Jrcrt Jyltt:
VII
256 STUDIES IN ANCIENT GREEK SOCIETY
being taken by the king masked in a bull’s head. 32 As son, he
is the Zeus whom Rhea entrusted to Amaltheia; as consort, he
is the Iasion whom Demeter embraced in a ploughed field. 88
— rrr — At Eleusis too the sacred marriage
led to the birth of a divine child,
jIik ^ who appears in myth as Demeter’s
foster-son, Demophon or Triptole-
' mos. 84 The use of the plough was
LX x\. communicated to mankind by Trip-
_p \ ^ tolemos, who had acquired it from
Demeter. 35 Thus the appearance of
‘w the male in the myth coincides with
fy{ [\ his intrusion into agriculture (p. 42).
V 6 (/ \ Demeter was of all Greek goddesses
4 jw\ W\ J/f the most conservative. The only centre
Vfj [WxWV at which her cult developed along
JL k- j J.. ¥ . — >\JL . E - \ / / . new l{ neS was Eleusis, where it be-
hg. 32 . Mmoan pnest: relief ca me a panhellenic mystical religion;
from nossos but even there she did not completely
FIG. 33. Demeter and Triptolemos: Attic cup
lose her agrarian character, and elsewhere she maintained her
ban on the other sex. She was the only mother-goddess to
survive intact.
32 Apld. 3. 1. 2, 3. 15. 8; Cook Z 1. 464-96, 521-5.
33 Od, 5, 125-7. 34 Horn. H. 2. 223-49, Apld. 1. 5. 2.
33 Harrison PSGR 273, cf. Paus. 8. 4. 1.
VII
SOME MATRIARCHAL fc»EITIES
257
2. Athena
Thucydides says that the earliest settlements in Greece were
not established on the coast, like the- later ones, but in the
interior, where they were safer from pirates. 38 His statement
may be illustrated by comparing some of the oldest sites —
Thebes, Orchomenos, Athens, Mycenae — with the Caro-Le-
legian settlements at Hermione, Epidauros, and the Messenian
Pylos. He also says that in these early times the largest element
in the population was the Pelasgoi. 37
The usual site for one of these early villages was a natural
eminence, which, when it expanded into a town, became the
citadel or acropolis. Athens is an obvious example. All over
Greece — at Athens, Argos, Sparta, Troy, Pergamos, Smyrna,
Rhodes, and many other places — we find the acropolis con-
secrated to Athena. 38 As Aristeides puts it, 'Athena reigns
supreme over the summits of all cities'. 38 This of course is an
exaggeration. She was not ubiquitous. But the association was
so widespread as to form one of her distinctive features, arid
it must have arisen from the circumstances in which her
worship was diffused.
She had deep roots in the Peloponnese. At Aliphera (Arcadia)
there was a local legend of her birth. 40 In northern Elis there
was a stream called the Larisos, where she was worshipped as
Athena Larisaia.« She had the same title on die acropolis at
Argos, which had been known in early times as Larisa. 42 This
was a Pelasgian place-name (p. 172). At Athens her cult must
presumably date from the foundation of the city. Its original
inhabitants were Pelasgoi, and Kekrops, their king, was her
servant. This suggests that she was brought there by her wor-
shippers, the Pelasgoi. From what direction did they come?
The Boeotians preserved the memory of a town called Athens
3 ® Th. x. 7. 37 Th. 1. 3. 2.
38 Paus. 2. 24. 3. (Argos), 3. 17. 1. (Sparta), 11 6. 88 (Troy), SIG.
xoo 7. 40 (Pergamos), Str. 634 (Smyrna), Pib. 9. 27. 7 (Rhodes, Akragas),
Paus. 2. 29. 1 (Epidauros), 2. 32. 5 (Troizen), 3. 23. 10 (Epidauros Limera),
3. 26. 5 (Leuktra), 4, 34. 6 (Korone), 6. 21. 6 (Phrixa), 7. 20. 3 (Patrai),
8. 14. 4 (Pheneos), 10. 38. 5 (Amphissa), X. Hell. 3. 1. 21 (Skepsis), GDI.
345 (Thessalian Larisa).
38 Aristid. 1.15. 4 <> Paus. 8. 26. 6. « Paus. 7. 17. 5. 42 Paus. 2. 24. 3.
R
258 STUDIES IN ANCIENT GREEK SOCIETY VII
submerged by the floods that created Lake Kopais. According
to Strabo and Pausanias it dated from a time when Kekrops
had ruled over Boeotia, implying that the Attic Athens was
fig. 34. Athena: Attic vase
the older;* 8 but there is no independent support for this tra-
dition, while the motive for inventing it is obvious. The
Athenians did not like to think they owed anything to the
stupid Boeotians. Yet it appears they were indebted to them
48 St. B. ’A6fjvai, Str. 407, Paus. 9. 24. 2; Meyer GA 2. I.. 277-8.
VII
SOME MATRIARCHAL DEITIES
259
for Athena. Her epithets Tritogeneia and Alalkomeneis, re-
ferring to the tradition that she was bom on the banks of the
Triton near Alalkomenai, must be extremely old, because they
are already divested of their local significance in the Homeric
poems . 44 For these reasons we may agree with Meyer that
Athena reached Attica from Boeotia.
Entering Boeotia, we are led still further north. In historical
times she was the national goddess of the Boeotian League,
which worshipped her as Athena Itonia. The epithet points
to Itonos in southern Thessaly. There, under the same title,
she was the national goddess of the Thessalian League . 46
The Boeotian League went back to the Boiotoi, who occupied
Boeotia in the period of the Trojan War. The Thessaloi overran
Thessaly in the same period: it was they who drove the Boiotoi
into Boeotia. From this it is clear that the cult of Athena
Itonia reached Boeotia from southern Thessaly. But, since her
cults at Athens and elsewhere date from long before the
-Trojan War, we cannot suppose that she was introduced into
Thessaly by the Thessaloi or Boiotoi. They must have taken
her over from the people they found there, and since the
Pelasgoi are said to have been more numerous in Thessaly than
in any other part of Greece , 48 we may safely conclude that she
had belonged in the first place to them.
Along the southern shore of the Gulf of Malis, which
divides Thessaly from Boeotia, lie the territories of Lokris
Epiknemidia and Lokris Opountia, so called to distinguish
them from another settlement of the same people, Lokris
Ozolis, on the Gulf of Corinth. The people of Lokris Opountia
had a remarkable custom. Every year they used to send two girls
to Troy, where they were dedicated to the service of the
Trojan Athena . 47 It was explained as an expiation for the sin
of Aias, their leader at the Trojan War. During the sack of
the city he raped Priam's daughter, Kasandra, who was a
priestess of Apollo. This of course is an etiological invention,
in which the truth is inverted. The custom itself implies that
44 Paus. 9. 33. 7, IL 4. 415, 5. 908 etc.
48 Paus. 9. 34. 1, 10. x. 10, Str. 41 x. Th. 1. 12. 3.
48 Str. 220-1.
47 Timae. 66, Plb. 12. 5. 6; Wilhelm LM, Kretschmer H 256-7.
z6o STUDIES IN ANCIENT GREEK SOCIETY'’ VII
the Locrian cult of Athena was an offshoot of the Trojan.
In the Iliad Athena is the patron goddess of Troy, residing in
her temple on the acropolis. But her priestess, Theano, is not a
Trojan at all. She is a daughter of Kisseus, who dwelt in
Thrace. Strabo says he dwelt in that part of Thrace known later
as Macedonia — in the peninsula of Chalkidike, where there
was a. Mount Kissos and at one time a town of the same name,
absorbed in the later Thessalonike (Salonika). 48 This region
had been inhabited by Pelasgoi, whose language survived there
(p. 1 71). Kisseus also appears as the eponym of the Kissioi, a
tribe located near Sousa in Lower Mesopotamia. 49 Sousa is a
far cry from Troy, but the House of Priam had oriental con-
nections. One of the Trojan allies was Memnon, son of
Tithonos, the founder of Sousa, whose acropolis was known as
the Memnonion; Tithonos was a brother of Priam, and his
wife was named Kissia. 60 Hecuba herself is described in the
post-Homeric tradition as a daughter of Kisseus. 61 It seems
then that there were two branches of the Kissioi, east and west,
the latter providing the dynasty of the Homeric Troy.
Memnon was sent to Priam’s assistance by one Teutamos,
who is described vaguely as a king of Assyria or Asia. 62 The
proper names in -amos have been investigated by Kretschmer,
who has shown that they are characteristically Anatolian. 68
The native name for the Pelasgian Hermes of Imbros (p. 172)
was Imbramos. 64 The Pelasgian chiefs allied to Priam — they
48 II. 11. 222-4, Str. 330. 24, cf. Lyc. 1232 sch.
49 Hdt. 3. 91. 4, 5. 49. 7, 7. 62. 2, Str. 728. The Kissioi of Sousa
were associated with die Medoi (Hdt. 7. 62. 2, 2. 86. 1, 2. 10. i, Plb.
5 * 79 * 7 ) who were traditionally connected with the Caucasus and had at
one time been known as Arioi (Hdt. 7. 62. 1); the Kissioi are coupled with
the Arioi in A. C. 422. The Arioi (Areioi) can be traced further east beyond
Hyrcania, where they had a town called Sousia: Arr. An. 3. 25. 1.
60 Od. 4. 188-9, II. 11. 1, Str. 728, Simon. 2 7 Bergk, ll. 20. 237,
Hdt. 5. 49, 53 _ 4 > A. Per. 123, fr. 405, D.S. 2, 22. This Kisseus was
sometimes identified as the Homeric Kisseus (E. Hec. 3 sch. AM) but
according to Philochoros ( FHG . 4. 648) he was the eponym of a 'Phrygian'
village or clan. Priam's mother is given in one version as Plakia (Apld. 3. 12.
3), which was a Pelasgian setdement (p. 171).
61 E. Hec. 3, Verg. A. 7. 320, 10. 705.
62 D.S. 2. 22, Io. Ant. 24=JFifG. 4. 550.
68 Kretschmer EGGS 325. 64 St. B. 'Ipppos, Head 261.
VII
SOME MATRIARCHAL DEITIES
26l
came from the Trojan Larisa — were grandsons of Teutamos. 88
The Thessalian Larisa had once been ruled by a Pelasgian
king named Teutamidas or Teutamias. 88 And Priam’s own
name (Priamos) belongs to the same class. All this hangs to-
gether. If the Kissioi were Pelasgian, we look for their ori ginal
home in the Caucasus (p. 173); and there we find a settlement
called Kissa, 87 the modem Kisseh, on the coast between
Trebizond and Batum. And if their two branches migrated in
contrary directions, that explains the myth of Tithonos which
the western branch brought into the TEgean: Tithonos was
carried off by the Dawn. 88 Were these Kissioi of Troy and
Macedonia the Pelasgians who, as suggested in Chapter V
(p. .193), introduced the culture characterised by ‘Minyan
ware’? I have little doubt that this and other questions affecting
the Pelasgian immigration could be answered by further
analysis of the literary data, mythographical and topographical;
but for the moment I am merely concerned to insist that
Athena was a goddess of the Pelasgoi.
Kekrops is described as a ‘son of the soil', that is, a Pelasgian
aborigine. His body ended in a snake's tail. 89 He had three
daughters — Aglauros or Agraulos, Herse, and Pandrosos. The
etymology of the first is not clear, but Herse, the ‘dew’, and
Pandrosos, ‘all-dewy’, emanate from the cult of the sacred
olive that grew in the Pandroseion adjoining the temple of
Athena Polias. 60 Once, when Athena visited Hephaistos in his
forge to ask him to manufacture some weapons for her, the
fire-god assaulted her, and in the struggle his semen fell on to
her leg. The disgusted goddess took a piece of wool (/ rion ) and
88 II. 2. 840-3, Str. 620; Leaf T 198-213.
88 Apld. 2. 4. 4. The original name of Theophrastos, a native of Lesbos,
was Tyrtamos, which Aristotle persuaded him to change because it was
so ugly: Str. 618.
87 Arr. Ini. 26. 8. According to Herzfeld 2, Kicmoi is derived from
Akkadian Kassu, Kassc, cf. Str. 522 KocroaToi from Aramaic qnssayt (mod.
Ba-qsa), all of which 'presuppose genuine Kas, from which the true plural
would be Kasip, attested by Gk. K&nnon’
88 Horn. H. 5. 218.
89 Apld. 3. 14. 1; Roscher LGRM 2. 1019: fig. 35.
88 Apld. 3. 14. 2, Paus. x. 2. 6, 1. 27. 2. Sacrifices were offered to
Pandrosos and Athena by the ffriPoi: IG. 2. 481.
262 STUDIES IN ANCIENT GREEK SOCIETY VII
wiped it off on to the ground ( 'chth&n ). From it sprang the snake-
child Erichthonios, whom Athena entrusted to the daughters
of - Kekrops ,(p. 222). 61 This story, in the form in which we
have it, is a singularly clumsy concession to the later Athenian
doctrine that their goddess was a virgin. The need for in-
venting it could scarcely have arisen unless the two deities had
HG. 35. Athena,, Erichthonios and Kekrops: Attic relief
once been united in some form of sacred marriage. The real
mother of the snake-child was the snake-goddess. And this
goddess, like the Minoan, was associated with a sacred olive,
tended by the daughters of the royal house. Her cult was
matriarchal.
Two important events were assigned to the reign of Kekrops,
One was the institution of matrimony (p. 142). The other was a
dispute between Athena and Poseidon for the possession of the
Acropolis. While Athena was planting her sacred olive in the
Pandroseion, Poseidon produced with a blow of his trident
61 Apld. 3. 14. 6.
VII SOME MATRIARCHAL DEITIES 263
the sacred well in the precincts of the Erechtheion. Kekrops
was asked to arbitrate, and gave his verdict for Athena . 82
Similar stories, always with Poseidon as one of the contestants,
were current at Troizen, Argos, and Corinth. He disputed
the possession of Troizen with Athena, and eventually agreed
to share it . 88 He was opposed at Argos by Hera, who defeated
HG. 36. Athena and tic daughters oj Kekrops: Attic vase
him . 84 At Corinth, where his rival was Helios, the sun-god,
the city itself was awarded to Helios, but Poseidon received
the Corinthian Isthmus, where the Isthmian Games were
founded in his honour . 68 In another tradition the Games were
founded in memory of Melikertes, whose mother, Ino, was a
daughter -of Kadmos . 86 Megareus, the eponym of Megara,
which lies on the Isthmus, was a son of Poseidon from On-
chestos, near the Minyan Orchomenos . 67 These traditions
were brought from Boeotia, not necessarily by the Kadmeioi or
Minyai, but by people who had been in contact with them.
82 Hdt. 8. 55, Apld. 3. 14. 1. The well, which lay inside the building
and was probably an artificial cistern, is said to have contained sea-water
(Paus. 1. 26. 5). It may be compared with the Babylonian apsu (S. H.
Langdon in CAH 1. 399, S. Smith 8) and the x«varnfo»ov or dii\aaaa in
the lepfiv of the modern Greek church: this is a small subterranean piscina
into which is thrown water that has been used for baptism and other ablutions,
i.e. the KaWppcrra (Antoniadis 10).
as Paus. 2. 30. 6. 64 Paus. 2. 15. 5» 22. 4’
86 Paus. 2. 1. 6. 66 Apld. 3. 4. 3.
87 Paus. X. 39. 5, Apld. 3. 15. 8. The Homeric Hisa, assigned to the
Boiotoi (IL 2. 508) is perhaps Nisaia, die port of Megara: Allen HCS 57.
VII
264 STUDIES, IN ANCIENT GREEK SOCIETY
In Thessaly we have met Poseidon as the ancestor god of the
Tyroidai (p. 191), and still further north he had an ancient
cult at Petra near the mouth of the Peneios (p. 201). 68
Poseidon came from the north. Who brought him to Attica
and the Peloponnese? Not the Tyroidai, who did not reach
Attica till after the Dorian conquest. Nor the Boiotoi, who
also claimed descent from him, because, though they penetrated
to the Peloponnese, they have left no trace in Attica. There
remain the Lapithai, whose presence is recorded in Attica,
Corinthia, and the northern Peloponnese. 69 The Attic
Peirithoidai were Lapithai, and Peirithoos was the companion
of Theseus, the Athenian national hero. This saga, in its
present form, is probably no older than the latter part of the
sixth century, when the figure of Theseus was elaborated by
Athenian nationalism as a counterpart to the Dorian Herakles. 70
Before that he had been a local hero of Marathon, where the
Peirithoidai belonged, and in the Iliad we meet him as a
comrade-in-arms of Peirithoos and Kaineus in Thessaly. 71 In
origin he was a Thessalian Lapith. This solves our problem,
because Theseus was intimately associated, not only with
Poseidon, whom Pindar and Euripides describe as his father,
but with Troizen and the Isthmus. 78
68 pi. p. 4. 1 3 8.
69 Augeias of Elis and the Homeric chiefs of the Epeioi (II. 2. 6x5-24)
were Lapithai: D. S, 4. 69, Paus. 5. x. II.
70 Toepffer TP 30-46, P. Weiszacker in Roscher LGRM 3. 1761,
Herter 245, Schefold 65-7.
71 E. Held. 32-7, Hdt. 9. 73, Eph. 37, II. 1. 264-8, Hes. Sc. 178-
82. The Philaidai of Brauron (p, 121) had Lapith connections. The
mother of Philaios was a daughter of the Lapith Koronos (St. B. OiAatBai),
eponym of Koroneia in Thessaly (St. B. Kopcbveia); there was another
Koroneia, the modem Koroni, near Brauron (St. B. l.c.). Among the personal
names current in this clan, were Kypselos (Hdt. 6. 34; see p. 201) and
Thessalos (Plu. Per. 29). The latter also occurs among the Peisistratidai,
who may have got it by intermarriage, because Peisistratos came from
Philaidai (Plu. Sol. 10).
78 Pi. fr. 243, E. Hip. 887, 1167-9 etc., B. 16. 33-6. The 8th of the
month was sacred to Theseus and Poseidon: Plu. Thes. 36, Prod, ad
Hes.- Op. 788. Theseus was born at Troizen and deared the Isthmus of
highwaymen:. Apld. 2. 6. 3, 3. 16. 1. One of the early kings of Corinth was
Marathon (Paus. 2. 1. 1)— another Lapith connection; and further, a
Lapith origin may be assigned to the religious confederacy known as the
VII SOME MATRIARCHAL DEITIES 265
Poseidon has two further links with the Lapithai, in-
dependent of Theseus and bearing directly on the point at
issue. The first comes from Corinth. Potidaia, a Corinthian
colony on the Macedonian coast, bears his name, which was
presumably chosen by its founder, a son of the tyrant Perian-
dros. 7S Periandros was the son and successor of Kypselos, whose
story was told in Chapter V (p. 201). And Kypselos be- '
longed to the Lapith clan of the Kaineidai, who had settled
near Corinth at Petra. 7 * It may be inferred that they had come
from the Thessalian Petra, the seat of Poseidon Petraios. The
Corinthian Petra cannot be precisely located, but, if it was
east of the city, where the main part of Corinthia lay, it was
on the Isthmus. Poseidon’s intrusion at Corinth is thus ex-
plained. He was brought there by a branch of the Lapithai.
In the Erechtheion, which had belonged originally to
Athena (p. 116), there was an altar of Poseidon Erechtheus,
symbolising the fusion of the old snake hero with the new god.
Beside it was an altar of Boutes, and the walls were adorned
with portraits of Boutadai who had served there as priests. 76
Boutes was a son of Poseidon (p. 126), and he had a daughter,
Hippodameia, whose wedding with Peirithoos was the oc-
casion of the celebrated fight between the Centaurs and the
Lapithai. 78 These Boutadai were another branch of the
Lapithai.
In Chapter V it was suggested that the Lapithai were one
of the peoples that brought Greek speech into prehistoric
Thessaly (p. 197). Now, the Boutadai were the cream of
League of Kalauria. Its centre was the temple of Poseidon on the island of
that name, which belonged to Troizen (Eph. 59) and its other members were
Hermione, Epidauros, Nauplia, Prasiai, Aigina, Athens, and the Minyan
Orchomenos: Str. 374. Hence the sanctity of the number 8? Its Boeotian and
Thessalian connections appear also in the tradition that the island had once
been called Anthedon or Hypereia: Arist. aj>. Plu. IS. 29 ^d, cf. II. 2. 508,
734. Poseidon is said to have taken the island in exchange for Delos from
Leto or Apollo: Eph. 59, Paus. z. 33. z. This harmonises with the tradition
of Carian settlements at Troizen and Epidauros (p. 170).
73 Nic. Dam. 60.
74 Hdt. 5. 9ZP.
76 Paus. 1. 26. 5, Plu. M. 841-3.
7 « D.S. 4. 70, II. 1. 263 sch. V, Od. zi. 295-309, Paus. 5. 10. 8, Apld.
Ipit. 1. 21.
266 STUDIES IN ANCIENT GREEK SOCIETY VII
the Athenian aristocracy — clannish, old-fashioned, and re-
actionary (p. 108). The elevation of Theseus to the status of
national hero may well have been due to them. Earlier in their
career they had captured the cult of Athena Polias, which, in
the conditions of aristocratic rule, implies virtual control' of
the city. If the Athenians were 'hellenised Pelasgoi’ (p. 176),
the Boutadai must have had a hand in hellenising them. The
view that they came of a Greek-speaking stock is thus in-
trinsically probable.
When the Lapithai first settled in Attica, they were sur-
rounded by alien Pelasgoi, whose culture they assimilated and
adapted. But old memories died hard. They were kept alive
for generations by the struggle for the land. The aristocrats
were as proud of their non-Attic ancestry as they were con-
temptuous of the common people, the natives, mere 'sons of
the soil'. The result was that the democratic movement took
the form of a resurgence of these ‘sons of the soil’, who, as
Munro has well said, boldly proclaimed their humble origin
as ‘a democratic slogan’ and 'a protest against the dominance
of an alien nobility’.”
Having usurped the shrine of Erechtheus, the Boutadai
proceeded to affiliate their clan ancestor to the native dynasty
(p. 126). Herodotus, our earliest authority for the dispute
between Poseidon and Athena, tactfully refrains from expres-
sing an opinion on the merits of the case, but Apollodoros says
definitely that Poseidon was the first comer and that Athena
only succeeded by the false testimony of Kekrops. 78 This ‘son
of the soil’, then, was a perjurer. Such were the little tricks by
which this lordly family legitimised its past.
But what of our Pelasgian Athena? She has been forced to
come to terms. She is still mistress of the Acropolis, but only
at a price. What that price was can be ascertained from another
version of the story. In this version the dispute is settled by a
democratic vote of the Athenian people — more democratic, in
fact, than any taken under the democracy. In the reign of
Kekrops, we are told, women as well as men had the right to
vote in the assembly. When the present dispute was sub-
mitted to them, the men voted for Poseidon, the women for
77 Munro 116. 78 Hdt. 8. 55, Apld. 3. 14. 1.
VII SOME MATRIARCHAL DEITIES 267
Athena, and the women got a majority of one. The goddess
was thus confirmed in possession, but the men retaliated. They
excluded the women for ever from the assembly, denied them
the tide of Athenians, and prohibited the practice of naming
children after the mother. 7 ® The women of this story are the
matriarchal Pelasgoi, the men are the patriarchal immigrants.
The conflict of cults coincides with the introduction of patri-
lineal succession, the disfranchisement of the women, and the
transition from group-marriage to monogamy. The myth ex-
presses as clearly as a myth can the unity of all human rela-
tionships — economic, political, social, sexual.
Changes so far-reaching must have taken a long time. Their
effect on Athena must have been equally gradual. It cannot be
followed in detail, but we are all acquainted with the final out-
come. With her serpent and sacred olive, and her girl priestesses
named after the dew, who carry her sacra underground and
play ball in her honour, the prehistoric Athena is hardly
distinguishable from the Minoan mother-goddess, whom we
see on signets and intaglios descending to her votaries as they
dance among the lilies, pluck her fruit, and twine her serpents
in their hair. These features were too deep to be eradicated,
but they were overlaid and reinterpreted. She had never been
married, because in Pelasgian times there had been no mar-
riage, but in the new age this is taken to mean that she prefers
virginity . 80 She had never had a mother, because as mother-
goddess she was herself the embodiment of motherhood, but
now she becomes the favourite daughter of Zeus the Father,
from whose head she sprang fully armed . 81 She remains a
patron of weaving, pottery, and the arts , 88 but in addition and
above all she becomes a goddess of martial valour, forensic
eloquence, and seasoned, temperate judgment — the ideals of
the democracy . 88 Her new official aspect appears in all its
forbidding splendour in the colossal gold and ivory statue
70 Varr. ap. Aug. CD. 18. 9; see p. 14a.
80 A.E. 740.
81 Hes. Th. 929 k-m: fig. 37.
88 Horn. H. 5. 7-1 5, cf. II. 9. 390, 01 . 7. Iio-I, 20. 72. Hence her
title Ergane: Paus. 1. 24, 3, 3. 17. 41 8. 32. 4, 9. 26. 8.
83 G. Thomson AO I. 56.
FIG. 37. Poseidon and Hcpbaistos at tie birth of Athena: Attic vase
has become as masculine as her sex, determined by her origin,
permits. It only remains to add that beneath the shield at her
feet there lay curled up unobtrusively a little snake —
Erichthonios . 88
84 Max. Tyr. 14. 6. 88 Paus. 1. 24. 7, Hyg, Ast. 2. 13.
VII SOME MATRIARCHAL DEITIES 269
3. The Ephesian Artemis
The cult of Artemis at Ephesos was already ancient when the
Ionians settled there. 88 If, as is believed, it goes back to the
Hittites, it must be at least as old as the thirteenth century. 87
No other Greek cult reveals quite such an unbroken history.
For what follows I am indebted to Picard, who has reconstructed
it from the Hittite period down to the days when Demetrius
made a fortune out of his devotion to Diana of the Ephesians. 88
It was addressed originally to Leto. She was represented by a
wooden image, said to have been found in the swamps
of the Kaystros, 89 which was hung on a sacred tree.
The earliest shrine was simply a courtyard surrounding
the tree, beneath which stood a small altar. 90 This
was a cult of exactly the same type as those depicted
on the Minoan gems (pp. 25 1 -3). In the early
archaic period this simple structure expanded into a
characteristic Greek temple — a house for the goddess
and her statue. 01 The temple was reconstructed
several times until it became one of the largest in
the Greek world, served by a populous and highly-
organised community of priests and priestesses. 92 The
annual festival fell in spring and lasted a month. 93
It opened with public sacrifices and dances, which
were followed by athletic contests. 94 These differed
from others of the same nature — the Olympic Games,
for example — in that as late as the sixth century
women were permitted to watch them without
restriction. 96 The winners were enrolled in a sacred / rom£ P 9fIM
college. 08 The general character of the Ephesian goddess is
thus delineated by Picard:
88 Paus. 7. 2. 6. 87 Lethaby ETA. 88 Acts 19. 24-7*
89 Picard EC 13-4. 90 Ii. 18-9. 01 lb. 20-1. 02 lb. 28, 104.
03 CIG. 2954. The month was Artemision, which Picard 328 equates with
the Attic Thaxgelion, but it may have been the Attic Mounychion, like
the Delian Artemision and Rhodian Arcamitios: SIG. 974 n, 5.
94 Picard EC 332.
95 Th. 3. X04. 3. At Olympia the rule was that girls might watch the
Games but not married women (Paus. 5. 6. 7) excepting the priestess of
Demctcr Chamyne (Paus. 6. 20. q). 98 Picard EC 340.
HG. 38.
Eplesian
Artemis:
statuette
270 STUDIES IN ANCIENT GREEK SOCIETY VII
The whole realm of nature belonged to her. She presided over the spring
blossoms and the fertilisation of the soil. She reigned over the elements,
ruled the air and waters. She governed the life of beasts, taming the wild
ones and protecting the tame. By turns a benefactress and a dealer of death,
healer of sickness and goddess of health, she was also the guide of souls on
their journey beyond the grave. 97
In addition, and in defiance of
the Homeric poems, which por-
trayed her as a virgin huntress
who abjured male company, she
remained to the last a helper
of women in childbirth . 9 8
The sacred tree marked the
spot where she was born. Leto
had leant against it. when the
birth-pangs came upon her . 99
This was the kernel of the cult.
Among the temple remains have
been found several statuettes of
the kourotrSphos type — a woman
nursing an infant. The oldest
of them represent simply a
mother and child — Leto and
Artemis. But in some of the
later specimens there are two
children. 100 The infant daughter
has been joined by an infant
son. Artemis eventually took her
mother’s place, but the Ephesian
Apollo never grew up.
Some twenty miles north of
Ephesos, near Kolophon, Apollo had a sacred grove called Klaros.
On this site too the original cult had been addressed to the
mother, Leto, and here again she gave birth to a child, but in
97 Picard EC 377. 98 Apul. Met. 11. 2. 99 Tac. Ann. 3. 61.
199 Picard EC 455-6, 479-81. For other examples of the kourotrSplos
type see Hansen 69 (Thessalian II, ‘a woman seated on a four-legged
stool holding a baby in her arms'), Nilsson MMR. 261 (Mycenean, from
Aigina, four idols of a woman with a child and one with two children’).
Like other goddesses, Artemis was worshipped as Kourotrophos: Farnell
CGS 2. 5 77.
FIG. 39. Mother-goddess and twins:
Attic vase
VII
SOME MATRIARCHAL DEITIES 271
this case it was a son, who eventually became supreme. 1 * 1
Why was she succeeded at Ephesos by die daughter, at Klaros
by the son? The answer is given by Picard:
In general Klaros was more resistant than Ephesos to the East and its
traditions. ... A god like Apollo, the celestial ruler of a patriarchal society,
would naturally receive a more favourable reception at Kolophon. 102
The Ephesian Artemis preserved her matriarchal character.
She had many oriental features, but these were not simply due
to oriental influences. 103 Rather, in admitting those influences
she remained true to her origin. Founded by die Hittites at the
height of their power, her cult was already, when the Greeks
wrested it from the Carians and Leleges, proof against any
radical alteration. But, though unable to patxiarchalise the cult
itself, the Greeks did introduce one innovation which politic-
ally was decisive. The sacred colleges included priestesses as
well as priests, but there was a rule that no woman might
enter the inner shrine on pain of death. 104 The central adminis-
tration was thus secured under male control. When we re-
member that these Greeks married Carian women (p. 169), the
significance of this rule becomes apparent, and its peremptory
character is a tribute to the tenacity of the matriarchal
tradition.
The temple of Artemis Orthia at Sparta stood in the marshes
of the Eurotas. Excavations have brought to light a number
of female figurines. 100 The image was said to have been dis-
covered in a bed of withies, and the goddess was named
Lygodesma, ‘ withy-bound'. 1 00 The parallel with the image at
Ephesos is so clear as to suggest that the Spartan Artemis was
an offshoot of the Ephesian. The earliest inhabitants of Sparta,
we remember, were Leleges (p. 170). It was at this shrine
101 Picard EC 455-6. 102 ll. 457*
103 That was in some cases the immediate cause, as when the Magi secured
a place in the cult after the Persian conquest: Picard EC 130.
104 Artem. 4. 4. Similarly, it was a capital offence for married women to
be present at the Olympian Games: Paus. 5. 6. 7. It appears that a non-Greek
clement survived in the Ephesian cult, for Aristophanes speaks of the goddess
as being worshipped there by Lydian girls: Ar. Nu. 599-600.
123 Dawkins 145-62.
too Paus. 3. 16. 11. Another tradition was that Orestes brought it back from
Tauris: Paus. 3. 16. 17.
272 STUDIES IN ANCIENT GREEK SOCIETY VII
that the Spartan boys, when they came of age, were subjected
to that ordeal of flagellation which has made their national
name a byword of austerity. The ceremony was a test of en-
durance or trial of strength — a typical form of tribal initiation
(p. 48). In one feature only was it abnormal. The boys were
scourged in the presence of the priestess, who held the sacred
fig. 40. Artemis Ortbia: ivory from Sparta
image in her arms. 1 **’ An invariable rule of primitive initiation,
enforced by the severest sanctions, is the rigid exclusion of
the other sex. The Spartan ordeal had therefore been modified
in this vital particular. The presence of a priestess at a rite
performed by priests means, as we have learnt from the Khasis
(p. 154), that the priestess had once officiated. We have also
learnt that the withy (Ijgos) was one of the plants used for the
sake of its supposed effect on menstruation (p. 218). An ordeal
of flagellation described as very similar to the Spartan sur-
vived at Alea in Arcadia, and there it was performed on
Jo? Paus. 3. 16. 11.
VII
SOME MATRIARCHAL DEITIES
273
women. 108 The scourging of the Spartan boys was done under
the eyes of the priestess because it was derived from a rite in
which 'the novices had been girls and the priestess the
officiant. 1 00
None of the remains on this site .antedate the Dorian con-
quest. This means that the cult was established there by the
Dorian settlers, but not that it was indigenous to them. It has
always been the concern of conquering peoples to consolidate
their position by adopting the cults of die conquered. And as a
matter of fact we know where the Spartan Artemis had come
from. Her site at Sparta was called Limnaion (llmm ‘marsh’),
and the cult had been so named after the village of Limnai
on the Messenian frontier, where there was a shrine of Artemis
Limnatis, goddess 'of the marshes'. 110
The meaning of the tide Orthia or Orthosia — it occurs in
both forms — is unknown. All we can say is that Orthia was a
village in Elis, Orthosia a village in Caria. 111 It was not
peculiar to Sparta. There are nine recorded cults of Artemis
Orthia (Orthosia), all of which except two are in the Pelopon-
nese. 112 Further, from what has just been said it is dear that
Artemis Orthia and Artemis Limnatis were virtually the same
goddess. There are seven cults of Artemis Limnads (Limnaia),
all of them in the Peloponnese. 113 To these may be added
Artemis Stymphalia at the Arcadian lake of diat name, and
Artemis Alpheaia at Letrinoi (p. 223), which clearly belong to
108 Paus. 8. 23. 1. In some cults of Demetcr, perhaps the Thesmophoria,
the women whipped each other with a plant oiled pfipavrov (Hsch. r.v.).
The underlying motive of these flagellations is clear from a Nandi rite
of initiation, in which the novices are beaten with stinging-nettles on the
genitals: Hollis NLF 54.
100 The Artemis Agrotera of Agra near Athens was also connected with
initiation; the 6pripo: held races and processions in her honour: IG. 2. 467-71.
“o Str. 362, Paus. 3. 16. 7, 4. 4. 2, 4. 31. 3, Tac. Ann. 4. 43.
111 Paus. 5. 16. 6, Str. 650.
112 Pi. O. 3, 54 sch., cf. Paus. 5. 16. 6, Hsch. ’Op 9 ta (Elis, Arcadia),
Paus. 2. 24. 5 (Mt. Lykone near Argos), Famell CGS 2. 572 (Epidauros),
CTG. 1064 (Mcgara), Pi. O. 3. 54 sch. (Athens), Hdt. 4. 87. 2 (Byzantium).
112 Paus. 2. 7. 6 (Sikyon), 7. 20. 7-8 (Patraf), 8. 5. xx (Tcgea), 4. 31. 3
(Kalamai), 3. 23. 10 (Epidauros Limcra), 3. 14. 2. (Sparta, E. of market-
place). In the last-mentioned cult she was also called Artemis Issora, as at
Teuthrone (Paus. 3. 25. 4).
S
CKM3IUV/ jn cnm
VII
SOME MATRIARCHAL DEITIES
^75
the same class. 114 . There was another Artemis Alpheaia at
Olympia . 116 Again, this goddess ‘of the marshes’ cannot have
been very different from Artemis Agrotera, the goddess ‘of die
wild’. This title occurs at nine sites, of which five are in die
Peloponnese . 116 Thus, out of twenty-four centres of Orthia-
Limnatis-Agrotera no less than eighteen are in the Pelopon-
nese. Of the remainder, two were at Athens; anodicr at
Byzantium, a colony from Megara; two more at Artemision
(Euboia) and Phanagoreia, both Ionian colonies; and the sixth
in Akarnania. These cults were characteristically Pelopon-
nesian, and that harmonises with the view diat they had been
introduced from Anatolia by Carians or Leleges . 117
Outside the Peloponnese Artemis was worshipped chiefly in
Boeotia. But her Boeotian tides were different. At Chaironcia
she was Soodina, ‘saviour from birthpangs ’; 118 at Thisbe she
was Soodina and Soteira, ‘saviour ’. 118 The latter recurs at
Megara, Troizen, and in Laconia and the southern Cyclades . 120
At Chaironcia, Thisbe, Thespiai, and Orchomenos we also find
an Artemis Eileithyia . 1 * 1 From this we may conclude that in
Boeotia the Carian goddess of childbirth merged with die
Minoan.
In Thessaly we meet Artemis Soteira at Magnesia, but the
typical Thessalian form of the goddess was Enodia, ‘of the
i 14 Paus. 8. 22. 7, 6. 22. 8.
118 Paus. 5. 14. 6, Str. 343.
118 X. Hell. 4. 2. 20 (Laconia), Paus. 5. 15. 8 (Olympia), 2. 29. 1 (Epi-
dauros), l. 41. 3 (Megara), 7. 26. 3 (Aigeira), r. 19. 6 (Agrai), Lolling
AAN 202 (Euboia), Supp. Epig. Gr. 1. 213 (Akarnania), CIG. 2117
(Phanagoreia).
U 7 In some cases from Crete. In several places she was identified with
the Cretan Britomartis-Diktynna (p. 255): Paus. 3. 14. 2 (Sparta), 10.
36. 5 (Phokis), E. Hip. 145, 1130 (Troizen), Homolle 23 (Delos).
118 IG. 7. 3407.
118 Schmidt 129, Latischcw 3 57, Famell CGS 2. 586. Throughout Boeotia
she was worshipped at marriage under the title Euklcia: PIu. Arist. 30.
128 Paus. 1. 40. 2, 1. 44. 4, 2. 3 r. 1, 3. 22. 12, C/C. 2481, Legrand
93. Also at Pellene, Megalopolis, Phigalia, and Athens: Paus. 7. 27. 3,
8. 30. 10, 8. 39. 5, Famell CGS 2. 586.
121 CIG. 1596, Famell CGS 2. 568, Schmidt 129, Ladschew 357. In
this form she was identified with Artemis Locheia: PIu. M. 6593, cf.
CIG. 1768, 3562, A. Su. 684-5.
276 STUDIES IN 'ANCIENT GREEK SOCIETY VII
crossways', 183 a local form of Artemis-Hekate, related perhaps
to the Thracian Brimo and Bendis. 138 Nowhere in Thessaly do
we find Orthia or Limnatis or Agrotera. Thus the northward
extension of these titles coincides with the limits of the Carian
domain (pp. 170-1).
4. The Brauronian Artemis
We have now to track this virgin huntress into one of, the
darkest corners of Greek religion.
The progenitor of the Arcadians was Arkas, the ‘bear-man*
(arktos). Shortly before his birth his mother, a companion of
Artemis, had been changed into a bear. 124 Her name was
Kallisto, Megisto, or Themisto. These were properly titles of
Artemis herself. 126 At Brauron, on the Attic coast, was a temple
of Artemis Brauronia. Here, before marriage, the girls, clad
in saffron, performed a bear dance. 126 Another incident in the
festival was the sacrifice of a goat. Once upon a time, after
killing a bear, the people had been afflicted by the goddess
with a plague, and in the hope of appeasing her one of them
sacrificed to her a goat which he had dressed up in his
daughter’s clothes. 127
The Arcadian myth, the expectant mother turned into a
bear, is explained by the Attic ritual, the bear dance of in-
tending brides. But in the ritual there are two details not
covered by the myth — the sacrifice of a goat and the pretended
sacrifice of a girl. The goat, we may suppose, was a substitute
for a bear. This would imply that the ritual was derived from
an earlier period, or from a foreign country, in which bears
were easier to come by than they were in historical Attica.
But what about the girl? We have heard already the story of the
Athenian children molested by Lemnian Pelasgoi (p. 175).
122 Stahlin 54, 71, 107, Sttpp. Epig. Gr. 3. 485.
128 Head 307-8, Lyc. 1176-80, Hsch. BevBts, App. EC. 4. 105, E.Ib 1048.
134 Apld. 3. 8. 2, Paus. 1/25. 1, 8. 3. 6-7, Erat. Cat . 1, Hyg. F. 155,
176-7. Similar totemic myths attached to the birth of Apollo (p. 156).
126 Muller PMW 73-6, Farnell CGS 2. 435. She was worshipped as
Kalliste at Athens and Trikolonoi: Paus. 1. 29. 2, 8. 35. 8.
i2e Ar. Ly. 645 sch., Harp. ApicreOaon.
127 Hsch. Bpocupwvla top-rfj, Eust. ad II . 331. 26, Suid. 'Eujiapis s!pi.
VII _ SOME MATRIARCHAL DEITIES 277
In another tradition the Pelasgoi are accused of raiding the
coast at Brauron, kidnapping Attic girls, and shipping them to
Lemnos. 128 What did they want with them there? The people
of Lemnos, we are told, worshipped a ‘great goddess’, to whom
girls were immolated. 129 Murder will out.
When the fleet of a thousand ships assembled at Aulis, it
was held up by storms, which the prophet interpreted to mean
that Artemis was angry and could only be placated by sacri-
ficing the king’s daughter. So Agamemnon prepared to
slaughter Iphigeneia, who was clad in saffron, but at the last
minute she was spirited away and replaced at the altar by a
hind or a bull or a bear. 130 She was carried overseas to Tauris,
the Crimea, whose king, named Thoas,'was in the habit of
sacrificing to Artemis every stranger that landed on his shores.
There she became the priestess of the goddess. Many years
afterwards her brother, Orestes, arrived, in exile for the
murder of his mother. The king handed him over for sacri-
fice, but, discovering his identity, Iphigeneia disguised him
in the sacred vestments of the goddess and on the pretext of
taking the image down to the sea to wash it embarked with
him on his ship and sailed safe home. 181 Let us return to
Lemnos. When the Lemnian women murdered their menfolk,
Hypsipyle spared her father (p. 175). His name was Thoas.
She rescued him by dressing him in the vestments of Dionysus
and conveying him to the shore. There they took ship and
sailed away to Tauris, where he became king. 132
The history of this bear-goddess is now plain. She belonged
to the Pelasgoi, who brought her to Arcadia from Attica, to
Attica from Lemnos, and ultimately from the far shores of the
Black Sea. That being so, she must have reached the Aegean by
128 Hdt. 4. 145, PIu. M. 247a, II. 1. 594 sch. A.
129 St. B. Afjuvo;, Phot, nsyi&nv 0e6v, Hsch. usyiXT) 0 e6s.
130 Prod. Chr. I. 2=Kinkel 19, A. A. 249, E. Li, 87-98, 358-60,
1541-89, Apld. Epit. 3. 21-3, Lyc. 186 sch.
131 E. IT. 28-41, Hdt. 4. 103. x, Apld. Epit. 3. 23. According to Euripides
the image was brought from Tauris to Athens and removed from there to Halai
near Brauron: E. IT. 89-91, 1446-67, cf. Paus. 1. 23. 7. In other versions
it is taken to Laodikeia (Cappadocia) or Sousa: Paus. 3. 16. 8, 8. 45. 3. All
these variants are in keeping with the view that the myth was Pelasgian.
132 Hyg. F. 15, cf. 120.
278 STUDIES IN ANCIENT GREEK SOCIETY VII
way of the Propontis. This was one of the districts in which
Pelasgian speech survived (p. 17 1), an d here we find a Bear
Mountain — the hill on which Kyzikos was built . 188 The
Caucasian origin of the Pelasgoi is thus confirmed, and other
scraps of evidence now fall into place. The Kaukones, whom
we traced in Elis, the Troad, and Paphlagonia (p. 171), bear
the Caucasian name; and in Chios, also occupied by Pelasgoi,
there was a village Kaukasa with a cult of Artemis Kaukasis —
the Caucasian Artemis. 184
The bear dance of Brauron was the initiation rite of a bear
clan in which one of the novices, incarnating the totem, was
put to death. Human sacrifice at initiation occurs sporadically
in modern tribes. 186 But the same goddess had other sacred
animals beside the bear. One of these was the bull, after which
she was named Tauro or Tauropolos. This title occurs in
Attica, Lemnos, and Cappadocia, 136 and of course it is implied
in the name of Tauris itself, her original home.
If the Artemis of Brauron was Pelasgian, we must consider
in what relationship she stood to Athena. We look for some
point of contact between the two cults. It turns up at Troy.
Those Locrian girls who were sent to serve the Trojan Athena
(p. 259) had first of all to undergo an ordeal. They were made
to run for their lives. If they managed to reach die sanctuary
without being caught, they became priestesses; if not, they
were sacrificed to Athena. 137 Again we recognise the ‘great
goddess' of Lemnos.
It appears, then, that, while the main body of Pelasgoi came
overland by Macedonia and Thessaly, another group, smaller
and perhaps later, reached Central Greece by sea from Lemnos
and the Troad. Athena belongs to the first movement, the
bear-and-bull-goddess to the second. Why then was the latter
named Artemis? The identification was due presumably to the
influence of the great goddess of Ephesos. And perhaps it was
133 Str. 575, Nic. Alex. 6-8.
134 Hdt. 5. 33. V, cf. SIG. 1014. 20, IG. 12. 5. 1078.
136 Webster 35.
138 Famell CGS 2. 569-70. Another of her Anatolian titles was Leuko-
phryene: SIG. 558. 12, 561. 26, Str. 647, Tac. Ann. 3. 62, cf. Paus.
1. 26. 4, 3. 18. 9. Leukophrys was an old name for Tenedos: Str. 604.
137 Lyc. 1141 sch.
• VII
SOME MATRIARCHAL DEITIES
279
first made in the Troad. From a fragment of the Kyprta, one of
the lost epics, we learn that, when the daughter of Chryses
was captured in the Troad by the Achasans, she was due to be
sacrificed to Artemis. 138 Apparently this Trojan Artemis was
only another form of the Trojan Athena, and the confusion is
explained by the fact that in this area the Pelasgian and Caro-
Lelegian domains overlap.
There were other affinities too. The Ephesian Artemis never
lost her maternal and lunar associations. 138 The Spartan
Artemis had a shrine just outside the town to which male in-
fants were brought by their nurses 148 — a variant of the custom
of showing the baby to the moon (p. 216). The Brauronian
Artemis was clad in vestments made from the clothes of
women in childbed, 141 and she bore the title Mounychia, 142
which undoubtedly refers to the moon. 143 There was a town
of this name near Athens, and here her festival fell on the six-
teenth of the month of Mounychion (April-May), 144 implying
that it was based on the old monthly observance of offering
cakes on the night after the full moon (p. 229). Artemis
Mounychia reappears at Pherai, Pygela, Kyzikos, and Plakia —
all within the Pelasgian domain. 146
The evolution of Artemis illustrates the truth that the Greek
deities are products of a complex process involving the fusion
of different cultures. This Pelasgian Artemis cannot be dis-
missed, any more than Athena, as non-Hellenic. She is pre-
Hellenic in the sense that the girls of Brauron had probably
been dandng their bear dance before a word of Greek was
spoken in that or other Attic villages, but she contributed all
the more largely for that reason to the mature Artemis, the
goddess who in the most renowned of all the tales of Hellas
demanded the blood of Agamemnon’s daughter; and she
138 Eust. ad II . I. 366. -
139 pjcard EC 368.
148 Ath. 139a, Hsch. KopvSctMicrrptm, Kvprrrol, cf, Plu. M. 657c.
141 E. IT, 1463-7.
142 Ar. Ly. 645 sch.
1 43 Mowuxlcc seems to stand for *povfowxIa, an epithet of the moon, as
• pqOvuxes for *povf6vuxes, cf. p. 222 n. 1 14.
144 piu, jvf, 3496
146 Call. Dian. 259, Str. 639, C1G. 3657, Lolling MK 155.
280 STUDIES IN ANCIENT GREEK SOCIETY VII
continued to be adored by the young women of Brauron down to
the day when the knell rang for this maiden mother to sur-
render her shrines to the Blessed Virgin. The origins of Hel- -
lenism cannot be relegated to the limbo of an impenetrable
past in the Balkan highlands or the steppes of the Ukraine.
They lie on Greek soil, just beneath the surface.
5. Hera
Hera deviates from type even more widely than Athena.
She was probably the first to shed her matriarchal character. At
Mycen®, the royal seat of Agamemnon in the Argive plain, she
became the national goddess of the Achaean federation that
laid siege to Troy and so was exalted at an early date as queen
of Olympus and wife of Zeus, the celestial ruler of the new
patriarchal world.
In the historical period she was worshipped, especially as
goddess of matrimony, in most parts of Greece, but her
Argive Heraion never lost its primacy. In the Iliad Mycenae,
Argos, and Sparta are the three cities she loves best. 14 ® The
Spartan cult of Hera Argeia was introduced from Argos. 147
The most northerly point at which a shrine of hers is men-
tioned is Pharygai, on the Gulf of Malis, and it was founded
by settlers from Argolis. 148 In Bceotia she had centres in most
cities, but her oldest cult in this region seems to have been
on Mount Kithairon at the head of the Corinthian Gulf. 14 ® She-
was worshipped all round the head of the Gulf— at Corinth,
Heraia, and Sikyon. 16 ® These territories had formed part of the
kingdom of Mycenae. 161 Excavation has proved that the cult at
Heraia was derived from the Argive Heraion, and tradition
said the same of her two cults at Sikyon. 168 Her temple at
146 II. 4. 50-2. 147 Paus. 3. 13. 8.
148 Str. 426. She figures in the myth of the Argonauts as Hera Pelasgis:
Apld. 1. 9. 8, A.R. 1. X4.
148 Paus. 9. 2. 7, 9. 9. 3.
188 Farnell CGS 1. 248.
161 See p. 394.
162 Payne P 22; Paus. 2. n. 1-2, Pi. N. 9. 30 sch. The kings of Sikyon
had been vassals of Agamemnon: Paus. 2. 6. 7. Her festival at Aigina
was introduced by settlers from Argos: Pi. P. 8. 1 1 3 sch.
VII
SOME MATRIARCHAL DEITIES
Z8Z STUDIES IN ANCIENT GREEK SOCIETY VII
A
Olympia, the oldest on the site, cannot be dissociated from
the tradition that the Games were founded by the Argive
Herakles . 188 At Athens she is inconspicuous, and had no shrine
on the Acropolis. In Euboia her myth and ritual are almost a
replica of the Argive . 154
If the focus of her worship on the mainland was the Argive
Heraion, it follows almost of necessity that she must have
reached the Argive plain from overseas. Her oldest image in the
Heraion, made of pear-wood, had been brought there from
Tiryns . 1 66 Tiryns was only a couple of miles from Nauplia,
where there was a cult of Hera Parthenos , 1 86 and Nauplia, with
its fine natural harbour, must have been the principal port of
call for Minoan traders. There is another good harbour at
Hermione, and here too was a cult of Hera Parthenos, with a
tradition that this was where Zeus and Hera landed when
they reached Greece from Crete. 15 ’
In the ^Egean there is only one centre with any claim to
challenge the Argive Heraion. The worship of Hera at Samos
was of acknowledged antiquity, and her temple there was even
larger than that of the Ephesian Artemis . 158 Her image was
said to have come from Argos, but the Samians denied this
and insisted that she was born under the withy-tree in the
sanctuary . 159 Samos, like Hermione, was a Carian settlement,
and its old name had been Parthenia . 160 Thus the Samian Hera
was related to the Hera Parthenos of Hermione and Nauplia,
while the legend of her nativity suggests contact with the
Carian Artemis.
There is nothing to show that Hera originated in Anatolia,
158 Pi. O. io. 23-59. The Olympian Hera, like the Argive, wore a bridal
veil (G. W. Elderkin 424-5); the Olympian Heraia was founded by
Hippodameia, whose bones had been brought from Mideia (Paus. 5. x 6. 4,
6. 20. 7), and was probably held in the month of Parthenios (F. M. Corn-
ford in Harrison T 230); and the stream Parthenias near Olympia (Paus. 6.
21. 7) corresponds to the spring of Hera Parthenos at Nauplia (p. 2 85).
154 Famell CGS 1. 253; see p. 285. i 88 Paus. 2. 17. 5.
158 Paus. 2. 38. 2. is’ St. B. 'Epuicbv, cf. Theoc. 15. 64 sch,
188 Hdt. 3. 60. 1. 189 Paus. 7. 4. 4.
160 Str. 637. Imbrasos, the stream flowing past the sanctuary, is a Carian
name (SIG. 46. 57-9 'IpppdaoiSos . . . *l|ip&pcnSos) related probably to the
Pelasgian Imbros, Imbramos (p. 172 n. 100).
VII
SOME MATRIARCHAL DEITIES
283
and the local tradition at Hermione has already beckoned us td
Crete. There, at Knossos itself, she was worshipped with Zeus
in a sacred marriage, 181 which was doubtless a survival of the
Minoan palace cult (p. 255). A short distance from Knossos
is the Cave of Amnisos, where she gave birth to Eiieithyia. 1 ® 2
For these reasons we may be sure that Hera is descended from
some form or aspect of die Minoan mother-goddess.
The sacred marriage was one of the most widespread fea-
tures of her worship. In her cult at Plataiai an effigy draped as
a bride was escorted to the top of Mount Kithairon. 1 ® 3 At
Athens there was an annual feast celebrating her union with
Zeus. 164 In Euboia the nuptials were located on Mount Oche. 16B
At Samos she was again represented by an effigy in bridal
costume. 166 At Nauplia annual mysteries were enacted at the
161 D.S. 5. 72; G. W. Eiderkin 424-5.
102 Paus. 1. 1 8. 5. She was worshipped as Eiieithyia at Athens and Argos
Roscher LGRM 1. 2091, Hsch. Eltaeuiay.
las p aus , j), 3, 3-9.
104 Phot, ltpij y&jjoj.
166 St. B. KApwtos.
166 Aug. CD. 6. 7, Lact. Inst. 1, 17. Pre-nuptial intercourse was per-
mitted in Samos, and there was a local myth of Zeus and Hera uniting
in secret: S. 14. 296 sch. A.
VII
284 STUDIES IN ANCIENT GREEK SOCIETY
Spring in which she bathed after marriage to renew her vir-
ginity. 16 ’ In this way the local Hera Parthenos was reconciled
with the official wife 'of Zeus. At Hermione the bridegroom
is said to have approached her disguised as a cuckoo — a
reminiscence of the bird epiphanies characteristic of the
Minoan goddess. 168
The sacred marriage was sometimes represented as a union
of bull and cow. That is the meaning of the myth of Io, who,
as Jane Harrison and Farnell for once agreed, is Hera’s double. 169
Io was a priestess of Hera. It was her father who set up-the old
pear-wood image at Tiryns. 170 Zeus fell in love with her and
forced her father to drive her out of house and home into the
water-meadows of Lerna. There she was turned into a~cow and
put to grass under the hundred eyes of the herdsman Argos,
who wore a bull’s hide. Then, pursued either by Zeus or his
jealous queen, she wandered all over the world, till at last she
came to the Egyptian Delta. With a touch of his hand Zeus .
restored her to her right shape and mind, and by the same
touch she conceived a son, Epaphos. After many generations
Danaos, a descendant of Epaphos, set sail from Egypt with
his daughters, landed at Nauplia, and settled at Argos in his
ancestral home. 171
'Such is the story as Aischylus tells it. We see at once that
Io has something in common with the daughters of Proitos
(pp. 226-8). In Egypt she was identified with Isis, whose sacred
animal was the cow. 178 How old this part of the story was is un-
certain, but there were other versions in which Egypt does not
figure at all. In Euboia it was said that she gave birth to
167 Paus. 2. 38. 2.
168 Paus. 2. 36. 1-2; Nilsson MMR. 285-94. In the original form, we
may suppose, she was approached by the cuckoo as such — a myth of
parthenogenesis (p. 243 n. 217, p. 287 n. 182).
169 Farnell CGS 1. 182, Harrison PHW 74-8.
179 Apld. 2. 1. 3, Paus. 2. 17. 5, Plu. Dacd. 10; Roscher LGRM 3. 1754.
171 A. Pr. 672-709, 733-61, 816-41, 872-902, Su. 1-18, 305, Paus.
4. 35. 2, Apld. 2. 1. 2.
172 Apld. 2. 1. 3, D.S. I. 24-8, Hdt. 2. 41; see p. 379. Just as her
voyage to Egypt was influenced by her association with Isis, so her crossing
of the Bosporos rests on a confusion with the cow cult of N. W. Anatolia:
Arr. fir. 35.
VII ' SOME MATRIARCHAL DEITIES 285
Epaphos in a cave on the shore near Karystos . 173 The pre-
historic inhabitants of Euboia were the Abantes, who settled
there under the leadership of Abas, an early king of Argos . 174
The name of the island — the isle 'of fair oxen* — is also signi-
ficant. Not only is it suggestive of Io, but it reminds us that
the Argive Heraion stood on the lower slopes of a mountain
called Euboia. It is said to have been so called after Hera’s
nurse . 176 This means that it was originally an epithet of Hera
herself. So Io bore her child on the hillside overlooking the
temple in which she served. The myth is thus reduced to the
initiation of a girl as a priestess who impersonated Hera in a
sacred marriage, the male part being taken by a priest got up
as a bull . 176 And in this form it corresponds
exactly to the myth of the Minotaur.
Pasiphae, the wife of Minos, fell in love
with a bull, and the craftsman Daidalos
constructed a hollow effigy of a cow, which
was then covered by die bull with her
inside it . 177 The offspring of this ingenious
union was the Minotaur, a man with a bull's
head. He is the counterpart of Epaphos,
whom ^Eschylus describes as- a heifer . 178 nG * 4 Z - The Minotaur:
Zeus and Hera were worshipped every- nossos
where as patrons of wedlock, the Olympian couple whose joint
benedictionwasbestowedon the lawful union of man and wife . 179
Farnell argued that this aspect of the two deities was so ancient
as to defy further analysis . 180 It is true of course that
173 St. B. ’Apcnrrfe, KApwros. Bull, cow, and calf appear on coins
of Katystos: Head 357. Hera retired there after a quarrel with Zeus:
Paus. 9. 3. 1. *
174 Pi. P. 8. 73 sch.
176 Paus. z. 17. 1.
178 Cook Z X. 464-96.
177 D.S. 4. 77, Clem. Pr. 4. 51.
178 A. Su. 41.
179 A. E. 214, Ar. Tb. 973-6 et sch„ Suid. TtXela, Poll. 3. 38, PPG. 2. 57.
180 Farnell CGS 1. 199-201. On the strength of this hypothesis he
suggested that the Sirens in the hands of the Argive Hera, which are a
variant of the Horai or Charites (p. 339), 'may simply denote the fascination
of married life' (x. 184).
286 STUDIES IN ANCIENT GREEK SOCIETY VII
as a form of the sacred marriage' it was essentially nothing
more than a ritualisation of the sexual act, which man .has
inherited from the animals. But of all its forms the matri-
monial was the latest. It is only by a convenient licence
that we speak of it as a ‘marriage* at all. The Greek sacred
marriage certainly goes back into the Mycenean period, and
beyond it, but not in the form of a union between Zeus and
Hera.
If Hera was descended from the Minoan mother-goddess,
Zeus cannot have been her original partner, because he is the
one member of the Greek pantheon of whom we can say de-
finitely that his name is Indo-European. He was introduced no
doubt at a very early date, but he must have taken some time
to establish himself. It seems probable that he owed his rise
to power to the Achasans, most of whom traced their pedigrees
to him, and, as we shall see in Chapter XII, the Achaeans
belonged to the Late Mycenean period, when the matriarchal
structure of JE gean society was undermined. This revolution
in the real world precipitated an upheaval in the world of
ideas. The old matriarchal myths were subverted. They did
not die out, but they were adapted and distorted almost out
of recognition. It is to this period that the marriage of Zeus and
Hera must be assigned.
If Zeus and Hera had always been the ideal matrimonial
couple, we should at least expect to find their union blessed
with offspring. 181 But we do not. Zeus has hundreds of children,
but Hera is not their mother. Hera has several, but Zeus is not
their father. Nor can their married life be described as
exemplary. In the Iliad their conjugal squabbles are an unfailing
source of laughter. From every point of view this ill-assorted
Olympian family is a palpable fabrication. Athena is said to
have sprung from Zeus’s head, but she had once been a
typical mother-goddess, who is by definition fatherless.
Artemis and Apollo are said to have been begotten by Zeus,
but the early shrines of Ephesos and Klaros knew only of a
mother. Ares and Hephaistos were sons of Hera before Zeus
was claimed as their father, but originally, since the one was a
Thracian and the other a Pelasgian, they can have had nothing
m Cook WWZ.
VII SOME MATRIARCHAL DEITIES 287
to do with either. 182 With the sole exception of Eileithyia, who,
since both can be traced to Knossos, seems to have a genuine
claim on her reputed mother, all these children are supposi-
titious. As Herodotus remarked, it was Homer and Hesiod
who made the Greek theogony: 183 that is to say, it was a product
of the epic tradition, which has its roots in the Mycenean
period.
As a form of the Minoan mother-goddess, Hera must have
had a male consort (p. 255). Who was he?
Herakles and Iphitos were twins, one divine, the other
mortal. 184 This was the starting-point of the Herakles saga. It
corresponds to the widespread practice of killing one of a pair of
twins, rendered necessary by the difficulty of rearing them 186
and excused by the belief that the one killed became immortal. 188
Herakles was born at Thebes, but his mother was a native of
the Argive plain, and the Argive plain was the centre of his
exploits. 187 His saga was thus located in the two main areas of
Mycenean culture, 188 and this implies that it was of Minoan
origin. One of its most remarkable features is the hero’s re-
lationship with the goddess of his mother’s birthplace. It was
she who cheated him of his heritage while he was still in the
womb and sent serpents to strangle him as soon as he saw the
light. 189 It was she who maddened him to murder his wife and
children, incited the Amazons to take arms against him, and,
when he returned from the ends of the earth with the cattle
382 Hera gave birth to Ares without the help of Zeus after touching a
flower: Ov. F. 5. 229-56. As Cook observes, ‘we sink here to the same
primitive stratum of ideas as that which ascribed the birth of Hebe to
a lettuce* (WWZ 367): Myth. Vat. 1. 204. For similar parthenogenetic
myths among the ancient Irish see Chadwick GL I. 2x6, and cf. Roscoe B
(1911) 48: ‘Women found to be with child unexpectedly might affirm that
some flower falling from a plantain which they were digging had caused
them to become pregnant.’ See further Frazer FOT 2. 372.
183 Hdt. 2. 53. 2.
i« 4 Hes. Sc. 48-52.
las Meek 357.
iso Frazer GB-MA 1. 267-9.
187 Apld. 2. 4. 6, 2. 5. x.
188 Nilsson MOGM 207.
189 II. 19. 95-133, Pi. N. 1. 33-40, Apld. 2, 4. 5-8.
288 STUDIES IN ANCIENT GREEK SOCIETY VII
of Geryon, sent a gadfly across his path which scattered the
cattle far and wide . 160 She was his implacable enemy from first
to last.
It is a recognised principle of mythological analysis that,
when the proper relationship between two concepts has been
disturbed, it is liable to be converted into its opposite. Hera’s
hostility to Herakles 'protests too much’, and, when we turn
from the literary version of the saga to local traditions, we
find memories of something quite different. She had a shrine
in Sparta which he built out of gratitude for the help she
had given him in his fight with Hippokoon . 191 When he
saw her struggling with the giant Porphyrion, he shot her
assailant dead . 192 When he journeyed to the Garden of the
Hesperides, she was there to greet him, and, when he returned
home with the Golden Apples, she was again ready with a
welcome for him . 193 In these traditions he is her partner and
assistant.
The Greeks tried to resolve this embarrassing contradiction
by saying that there were two heroes of the same name — the
hairy-armed stalwart from Argos and a mild-eyed young man
from Crete, the latter being the older . 194 At Megalopolis there
was a statuary group including Demeter and her daughter
with this Cretan Herakles at her side . 195 At Mykalessos the
same Herakles served Demeter as sacristan . 196 According to the
literary tradition the founder of the Olympic Games was the
Argive Herakles, but the local priests, who ought to have
196 E. HF. 843-73, Apld. 2. 4. 12, 2. 5. 9-10.
191 Paus. 3. 15. 9.
192 Apld. 1. 6. 2.
19 3 Gruppe 460-1. These apples, which had been grown for Hera's
marriage (Ath. 83c), were probably pomegranates or quinces (p. 219).
Apples and quinces were, and are still, used as offerings of love or marriage:
Theoc. 2. X20, 3. 10, 5. 88, Verg. E. 3. 71, Claud. EP. 8, Polites E no. 138,
cf. Skirnismal 19: ‘Eleven apples all of gold here will I give thee, Gerth.’
Attic brides were recommended to eat a quince before lying down with the
bridegroom: Plu. M. I38d.
104 Hdt. 2. 43-4, Paus. 9. 27. 6-8.
196 p aus . 8. 31. 3. Herakles was associated with Demeter Eleusinia on
Mount Taygetos: Paus. 3. 20. 5.
196 Paus. 9. 19. 5, 9. 27. 8.
VII
SOME MATRIARCHAL DEITIES
289
known, said it was the other one from Crete. 1 ” When Dio
Chrysostom was on a country ramble near Olympia, he came
across a wayside shrine of Herakles with an old peasant woman
sitting beside it. In reply to his questions she explained in
broad Doric that she was the keeper of the shrine with the gift
of prophecy from the Mother of Gods. The local farmers used
to consult her about the welfare of their herds .and crops . 188
Joint cults of Herakles and the Mother seem to have been
common in the country districts . 199 In these out-of-the-way
parts the peasantry continued to worship the hero in his
ancient aspect. The only concession they had made to the
official view was to transfer him from the Olympian wife of
Zeus, who meant very little to them, to a goddess who had
preserved her homely, matriarchal character.
Then there is the name itself. Throughout his history,
whether Hera’s enemy or Demeter’s friend, the hero was
known by a name which means 'called after Hera’. It has been
strangely misconstrued. After analysing at length the Argive
and Theban Herakles, without saying a word about the
Cretan, Nilsson states that 'the name Herakles is the starting-
point for the role of Hera in the Herakles saga '.* 00 On this
Hypothesis, the original choice of the hero’s name is an
accident; his subsequent association with the goddess is.
another accident; and their hostility remains a mystery. We
do not deserve to solve the problem if we throw away the clue.
His name cries aloud to us that he is the mother-goddess’s
male partner, typifying the status of the sexes in a society in
which the son is named after his mother.
187 p£ 0. xo. 23-59, Paus. 5. 7. 6-7. Here again he was probably associated
with Demeter, whose cult epithet at Olympia was XojjCwj (Paus. 6. 20. 9,
6. zi. 1), i,e. XoiKxieOvi) ( 11 . 16. 235), in allusion to the tradition that he
and his companions used to sleep on beds of olive leaves (Paus. 5. 7. 7):
see my AA 115. The two versions may be reconciled on the hypothesis that
the cult underlying the Games was founded from Mycenas at a time when
the Argive Herakles retained his matriarchal character.
1 »® D.Chr. 1. 6xr.
wo Famell GHC 129.
200 Nilsson MOGM zii. The Greeks were quite clear about it: Herakles
was named after Hera either because he performed his labours at her
instigation or because he had saved her life in the Battle of the Giants:
Pi. fr. 291, IM. Neitos, cf. Kretschmer MN 122.
T
VII •
- ffawr~
290 STUDIES IN ANCIENT GREEK SOCIETY
The double axe was a symbol of the lightning (p. 251). At
Mylasa, the Carian capital, there was a national cult of Zeus
Labrandeus, Zeus of die Double Axe. ‘Why is it', asks die
indefatigable Plutarch, ‘that the Carian Zeus is portrayed with
an axe in his hand instead of the sceptre or the thunderbolt}’
When Herakles killed the Queen of the Amazons, he stripped
her of her arms, among diem
an axe, which he presented
to Omphale, the Lydian
queen in whose service he
was engaged. From her it
was handed down as an heir-
loom to the last of , the
Herakleidai, who was slain
by Arselis of Caria. Arselis
took it to Mylasa and placed
it in the hand of Zeus. 201
Zeus Labrandeus got his axe
from Herakles, who got it
from the Hittites.
An early Etruscan funerary
monument represents a
warrior carrying a double axe
and wearing on his helmet
an enormous crest. 202 The
crest was a national charac-
teristic of the Lycians and
Carians, who are said to have
invented it. 203 Further, the
fig. 43 . Etruscan amour: stele from Vetulonia
Etruscan Herkle and Unial and the Roman Hercules and Juno
stood in exactly the same relationship as we have postulated
for the Greek Herakles and Hera. On an archaic Roman
bronze we see Jupiter introducing Hercules to Juno. His in-
tention is something more than a reconciliation. That is
proved by the male and female genitals lying at their
201 Plu. M. 3oif. This Lydian Herakles was the same as Sandas, the consort
of Kybebe or Kupapa (p. 512): O. Hofer in Roscher LGRM 4. 319-33.
202 R. S. Conway in CAH 4. 392: fig. 43.
203 Hdt. I. 171. 4; Hall CGBA 136.
fig. 44. Juno and Hercules: Etruscan bronze
the bridegroom untied on the nuptial couch, was called the nodus
Herculaneus . a °« This evidence. has been cited by Cook, whose
204 Cook WWZ 374, Roscher LGRM 1. 2259: fig. 44. An Etruscan mirror
representing Herakles and Hera, with Zeus and Hebe in the background,
carries an inscription hercle unial clan, which means 'Herakles son of Hera';
Cook WWZ 416.
aos Fest. 63.
292 STUDIES IN ANCIENT GREEK SOCIETY VII
conclusion is that 'when, at a very early date, the cult of
Herakles spread to Italy, the acknowledged partner of Herakles
was Hera '. 908 The extant data do not permit us to decide when
the Greek myth was recast, but in at least one local cult the
hero preserved his marital function. At Kos marriages were
solemnised in his temple, and food was offered to him as one
■ of the wedding guests . 207
That is as far as we can see clearly, but faint signs tempt us
still further into the past. The union of Herakles with Hera of
Knossos reminds us mat Demeter embraced Iasion in a Cretan
field . 208 The companions whom Herakles brought with him
from Crete to found the Olympic Games were Paionaios,
Epimedes, Idas, andIasios. 209 Thefirsttwo are merely eponyms
of primitive medicine; Idas is named after the Cretan Ida; and
Iasios is barely distinguishable from Iasion . 210 This gives us a
whole series of reduplications — Herakles-Iasion, Demeter-
Persephone, Hera-Eileithyia, Eileithyia-Eleusis. When we
are in a better position to investigate Hera and Demeter on
their native soil, we may be able to track both down to their
origin in the neolithic mother-goddess.
One more question. If Hera divorced Herakles to marry
Zeus, who was the original wife of Zeus? Aristotle informs us
that the earliest home of the Hellenes was the country round
Dodona , 211 where Zeus had a shrine of immemorial antiquity,
perhaps his oldest on Greek soil. It is here, if anywhere, that
we might hope to find his Indo-European aspect surviving free
of Aegean influences. At Dodona, we are told, Hera was
-called Dione ; 218 and Dione, or Dia, is simply the feminine of
Zeus (IE *dyeus ). In keeping with the status of the sexes in
each case, the patriarchal Indo-European goddess was named
after her master, just as . the matriarchal Minoan god was
named after his mistress: and the fusion of the two cultures in
patriarchal Greece was aptly symbolised in the marriage of the
matriarchal goddess to the patriarchal god.
soo Cook WWZ 375. 207 Paton76. 208 Oi. 5. 125-7. sob p aus . 5. 7. 6.
210 Picard PPD 357. 211 Arist. Mete. 1. 14. 21a Qd. 3. 91 sch.
VII
SOME MATRIARCHAL DEITIES
293
6. Apollo
This account of the subject is not intended to be exhaustive .
One major goddess, Aphrodite, has been omitted. I shall have
something to say about her in dealing with the Homeric Helen.
The present chapter will be concluded with some observations
on Apollo, designed to show how he evolved out of the
matriarchal worship of Artemis and Leto.
Some of Apollo's characteristics, such as his connection with
FIG. 45. Apollo ani Artemis : vase from Melos
the amber trade, point northwards into Central Europe. 213
These may be Indo-European. But in general his affinities lie
with S.W. Anatolia and Crete. This has been shown by
Nilsson.
Festivals of Apollo are comparatively rare on the mainland, and he has
everywhere usurped older festivals which did not originally belong to him.
... In contrast to all the other Greek gods, who preferred the time of full
moon, Apollo occupied the seventh day of the month, on which all his
festivals are celebrated. The agreement with the Babylonian slabattu is
complete and cannot be accidental. . . . His mother, Leto, originated in
S.W. Asia Minor. Personal names compounded with Leto occur only here —
an argument of the most convincing kind. Her name is connected by
213 Krappe AK. It appears that, though Anatolian, Apollo was not
indigenous to Lycia (Konig SX iz) and the name Apulunas, referring to a
god of gateways (Apollon Agyieus?) has recently been deciphered on a
Hittite altar: Nilsson GPR 79, cf. Laroche 80.
294 STUDIES IN, ANCIENT GREEK SOCIETY VII
philologists with the Carian word lada, 'woman'. Her cults in Greece are few
and their age uncertain; only in Crete is a festival attributed to her. 214
Nilsson's results can be carried a stage further in the light of
Picard's work at Klaros. We saw how the Madonna of Klaros
was supplanted by her son. A similar development may be
presumed for the Apollo of other Carian settlements, especially
Miletos and Delos. In historical times his cult at Delphi
became so influential as to dominate all the others, but even
at Delphi it was remembered that the first keepers of his
shrine were strangers from Crete . 218 If he reached Delphi from
Crete, we may be sure that he reached Crete from Anatolia.
When the Carian Apollo came to Delos, his mother, the
Woman, was still strong enough to secure a place there for
herself and her daughter . 216 But when he landed under Par-
nassos, he proclaimed himself simply ‘the son of Zeus ’. 217 At
Delphi, in myth and ritual alike, the mother and sister,
dropped out . 218 The Delphic Apollo is thus a faithful image of
the social changes that had created him. At Ephesos the divine
heritage passed from mother to daughter; at Klaros and Delos
from mother to son. At Delphi mother and daughter both
withdraw, leaving the Son invested with the authority of his
almighty Father — a figure so commanding that we almost
forget he began life as a baby in the arms of a neolithic
figurine.
214 Nilsson MMR 443-4, cf. PTR 366-7, Picard EC 458-9, 463.
With Leto ‘the woman’ cf. mu-al-li-ia-at (MOAitto) 'the woman who bears, ’
a title of the Babylonian goddess of childbirth: Langdon BEC 217.
216 Horn. H. 3. 475-80.
216 His birth was located there: Call. Del. 36-58, Simon. 26b Bergk.
The Ephesian origin of the Delian cult is indicated by the old name
of the island, Ortygia (Ath. 392 d), which was the name of the grove in
which he was said to have been bom at Ephesos (Tac. Ann. 3. 61); and its
original connection with childbirth and initiation, especially the initiation
of girls, appears from Hdt. 4. 34-5, Call. Del. 255-7, 296-306, Paus.
1. 18. 5, 8. 21. 3.
217 Horn. H. 3. 480
218 Farncll CGS 2. 465.
Part Three
COMMUNISM
And the land shall not be sold in perpetuity; for the
land is mine.
Leviticus
My -field was God's earth. Wherever I ploughed, there
was my field. Land was free. It was a thing no man
called his own. Labour was the only thing men called
their own.
TOLSTOY
vm
THE LAND
i. Beginnings of Private Property
It is characteristic of hunting peoples that the huntsman does
not appropriate his catch but brings it home to be distributed . 1
This rule corresponds to an economy in which, owing to the
low level of technique, production and consumption were alike
collective . 2 As labour becomes more productive, a man tends
to claim for himself and his immediate relatives the wealth he
has acquired with his own hands. This is the germ of private
property and the family, which ultimately transforms the
tribal system into the state. In its initial stages, however, it
develops within that system, and even strengthens it by in-
tensifying those co-operative functions on which, as we have
seen, the tribe depends. Clan is tied to clan in an intricate
network of reciprocal services, in which, animated by a spirit
of constructive emulation, they vie with one another for
prestige . 3 The man who has acquired a surplus of game or loot
signalises his success by inviting'another clan to feast with his
own. His invitation is a challenge, imposing on his rivals an
obligation to return it, if possible with interest, in order to
recover their prestige. * If the obligation cannot be met, it
may be commuted into some form of labour service. And so
the clans cease to be equal. Co-operation becomes competi-
tion. Meanwhile the same process is beginning to take effect
within the clan, which accordingly splits into families.
1 Spencer NTCA (1904) 609, NTNT 36, A - 52, 490, Howitt
NTSEA (1904) 756, Malinowski FAA 283-6, Bancroft 1. 1x8, 417, 506,
Rivers KSO 108, Williamson 3. 235, Wollaston 129, Smith and Dale x.
384, Hobhouse 244, Landtman 7, Buradkar 155.
2 W. E. Roth 96, 100, Mathew 87, Hollis NLF 24, cf. J. L. Myres
in CAH 1. 50.
3 Morgan AS 96, Spencer NTCA (1904) 164, Hubert 195, Landtman 70.
4 Bancroft 1. 192, 217, 2. 711, Roscoe B (1911) 6, Frazer TE 3. 262,
300-1, 342-4, 519, 545, Granet 165, 267, Gronbech 2. 8, 87, Hubert
54, 193-6.
298 STUDIES IN ANCIENT GREEK SOCIETY VIII
Among modem tribes these inherent tendencies towards the
growth of property have been sharply stimulated by capitalist
exploitation. They mark the extreme point to which individual
rights can develop within the tribal system; and consequently,
in arguing back to the prehistory of civilised peoples* we must
be prepared to find that common ownership persisted to a
higher stage. There are many indications, when we turn to our
own past, that one of the major factors was the adoption of a
pastoral economy. The Latin pecuttia, from pecus ‘cattle’, tells
its own tale, and it is supported by similar etymologies in
many other languages. 8 Game is perishable; land is immovable;
but livestock is easy to seize, divide, or exchange. Being neces-
sarily nomadic, pastoral tribes are quick to augment their
wealth by cattle-raids and war; and since warfare is waged by
the men, it reinforces the tendency inherent in this economy
(p. 42) for wealth to concentrate in their hands. These hardy,
restless tribes plunder one district after another, killing the
men and carrying off the women as chattels, until eventually
they settle permanently in an agricultural region and subject
the natives to regular tribute, which is the first step to reducing
them to serfdom. 0 Such was the origin of the Kassites who
overran Babylonia, the Hyksos kings of Egypt, and the
Ach$an pillagers of Minoan Crete. 7 The Indo-European
nomads possessed a further asset in the swiftest of all domestic-
able animals, the horse. Their success in extending their
speech so far was not due to any innate superiority but to a
peculiar combination of social and historical circumstances,
which gave them the opportunity to subdue and assimilate
the sedentary agricultural civilisations of the Near East.
8 Hcichelhcim I. 47.
0 Cf. Roscoe BB (1923) 6-9. The initial stage can be seen in Strabo's
account of the Massagetai and other Caucasian nomads, who secured over
the sedentary plainspcoplc the right to overrun and plunder their territory
at stated times of the year: Str. 51 1, cf. 3 1 1.
7 The speech of the Kassites, who conquered Babylonia c. 1710 B.c.
and introduced the horse, was largely Indo-European: Hall AHNE
199-203. Tire Hyksos or ‘shepherd kings’ who entered Egypt c. 1800 B.c.
included Anatolian and Indo-European elements, and the rapidity of their
conquest has been attributed to their use of the horse-and-chariot: ib.
212-3, Engbcrg 23, 41-50. This, derived probably from Egypt and Ana-
tolia, appears in Crete in Middle Minoan III: Hall CGBA 84-5.
VIII
THE LAND
299
War demands unitary leadership, and hence in these tribes
the kingship is militarised. 8 After a successful campaign the
king and his subordinate chiefs are rewarded with the lion's
share of the spoils, both chattels and land, and the wealth thus
accumulated promotes inequalities that shake the whole
fabric of society, beginning at the top.
2. The Problem of Ownership in Early Greece
In the Cambridge Ancient History, which lavishes a whole
chapter on the ‘famous victory’ of Marathon, the problem of
early Greek land-tenure is settled in one sentence:
The Greeks had long outlived the stage, if it ever existed, when land had
been owned in common by the' clan and private ownership was unknown. 8
Is it possible, then, we are prompted to ask, that private pro-
perty had existed ever since the enclosure of the Garden of
Eden? On that point the cautious writer does not commit
himself. It is enough to have pushed it back so far that its
origin can be comfortably ignored. This is hardly the way to
write history.
In the Iliad we read of
two men with measures in their hands quarrelling over boundaries in a com-
mon ploughland, contending for equal shares in a small space of ground. 18
What sort of tenure does this imply? Hardly the same as ours,
because the land is described as common. If we want to under-
stand it, we must study it in its context along with all the
other data bearing on the subject. This might seem to be
elementary commonsense. Yet here again our leading authori-
ties, usually so meticulous, become disconcertingly abrupt.
Listen to Nilsson, the greatest living Homeric archaeologist:
It is an old assumption that Homer mentions landed property as com-
munal and that this property was redivided from time to time, but the
passage adduced as evidence can be interpreted otherwise. It is uncertain
whether the word epixynos signifies ‘communal’; it may signify simply
‘common’, viz. ‘of disputed ownership’, and the quarrel may be one of the
quarrels concerning boundaries common among farmers. 11
8 See pp. 328-31. 8 F. E. Adcock in CAH 4. 42.
10 II. 12. 421-3. 11 Nilsson HM 242.
300 STUDIES IN ANCIENT GREEK SOCIETY VIII
WKat is meant by this subtle distinction between 'communal*
and 'common* and the even subtler equation 'common viz. of
disputed ownership* — these are questions that may try the
reader’s mother-wit as sorely as they have tried mine; and
even if he succeeds in answering them, he will still have to ask
himself why, if it is the ownership of the land that is in dis-
pute, the parties are engaged in dividing it into equal shares.
Further, the 'old assumption’ discarded in favour of this
cryptic hypothesis is the interpretation given half a century
ago by Esmein and .deduced from a comparative study of the
subject. I 2 The 'old assumption’ was consequently not an as-
sumption at all but a reasoned argument, which Nilsson has
replaced with an entirely unsupported assumption of his own —
that the passage can be interpreted out of hand in the light of
modern capitalist property relations. This, again, is not the
way to write history.
Why are bourgeois historians so shy of private property?
They were not always so. The earliest of them — Ferguson,
Millar, Adam Smith — were proud of it. They believed that
human progress depended on it, as indeed it did. These
writers anticipated Marx and Engels in recognising it as the
decisive factor in the growth of civilisation. They could not
fail to recognise it, because in their day the development of
capitalist property, for which they stood, was still being
obstructed by remnants of feudalism. How different the
bourgeois attitude to property was in those days may be judged
from some observations by Sir John Sinclair, an ardent ad-
vocate of the Enclosure Acts, in 1795:
The idea of having lands in common, it has' justly been remarked, is to be
derived from that barbarous state of society, when men were strangers to
any higher occupation than those of hunters or shepherds, or had only just
tasted of the advantages to be reaped from the cultivation of the soil. 1 3
In contrast to what we read in the Cambridge Ancient History,
this statement by an unlearned landlord in the days of ‘bad
King George* is scientifically correct. Of course, the bold asser-
tiveness of the old attitude and the evasive reticence of the new
both spring from die bourgeois interest in property. But the
12 Esmein PFPH.
is Hammond 12.
VIII
THE LAND
301
world has changed. Owing to the growth of the socialist move-
ment and more recently to the example of the Soviet Union
it is no longer feasible to dismiss communism as something
prehistoric, and so the subject has become taboo. It would be
superfluous to point out which attitude is the more conducive
to the discovery of truth.
Marxists are sometimes accused of distorting the facts to
fit their principles. The shoe is really on the other foot. It is
a habit of the bourgeoisie to charge their opponents with their
own delinquencies. The inductive method, which these em-
piricists profess, serves well enough for certain purposes, so
long as it is applied without restriction to the whole range of
relevant material, though even then it is inadequate; but when
it is confined, as in the present instance, to a small comer of
the field, which cannot be understood except in relation to
the whole, its effect is merely to preclude die possibility of
establishing general conclusions. In the eighteenth and nine-
teenth centuries the comparative method, without which
modem science would not exist, was applied by bourgeois
historians with magnificent results; but in recent times, with
far more material available, they have abandoned it. Con-
fronted with the growing power of socialism, they have re-
treated from one after another of the positions which their
predecessors gained. If private property came into being,
private property will pass away. ‘Ah, Faustus, now hast thou
but one bare hour to live.' If, on the other hand, its origins can
be pushed out of sight, we can shut our eyes a little longer to
the shadows that are creeping over it to-day. O lente, letite cur-
rite, noctis equi. And so the writing of history becomes more and
more introverted. 'It ceases to be a science and becomes an ‘art’.
The wilful blindness to which this attitude leads can be seen
in the remarks made by Toutain on the problem before us in
the Iliad:
Here we have a perfect picture of collective property, says Esmein.
Really, one must be die slave of a preconceived idea to interpret the scene in
this way. On the contrary, it seems to me that the attitude of the two
neighbours bears witness to the existence of private property and to the
stubbornness with which each fought for his own portion.* 4
* 4 Toutain 14.
302 STUDIES IN ANCIENT GREEK SOCIETY VIII
That is all — no argument) no reply to Esmein’s arguments.
Which is the slave? And this blank negative is combined with
a denunciation of the comparative method:
Because in some primitive peoples ownership of the soil has been col-
lective, that is no reason why the same system should have existed uniformly
in all primitive peoples. Those who draw this condusion forget that the
character of landownership cannot be independent of the nature of the soil
and the climate. ... In any case, a method which would, in such a matter,
draw condusions from one country to another is in my opinion thoroughly
dangerous, i 5
The particular form of land-tenure is determined in each case,
not only by soil and climate, but by the whole complex of
natural and social conditions. That is the proposition, which
Toutain refuses to face. On the other hand, it may be admitted
that there is a certain danger in drawing conclusions ‘in such a
matter* from one country to another on the continent .of
Europe in its present fluid state.
3. Primitive Land-tenure
It is time to enquire into the facts. This will not be easy.
The history of primitive land-tenure has yet to be written. In
Greece and elsewhere there are many problems still unsolved.
The most that can be attempted here is to outline the method
which, when pursued with more knowledge than I can com-
mand, will lead to their solution.
Let me begin by summarising the results obtained by Hob-
house, Wheeler, and Ginsberg from their statistical analysis of
the ethnological data:
Wc may express the whole tendency best by saying that the communal
principle predominates in the lower stages of culture and retains a small
preponderance among the pastoral peoples, and that private ownership tends
to increase in the higher agricultural stages, but partly in association with
the communal principle, partly by dependence on the chief, or in some instances
by something in the nature of feudal tenure. We seem in fact to get something
16 It. 12-3. Contrast Vinogradoff GM 18: ‘There seems to be hardly
anything more certain in the domain of archaic law than the theory that
the soil was originally owned by groups and not by individuals, and
that its individual appropriation is the result of a slow process of develop-
ment.'
VIII
THE LAND
303
of that ambiguity as between signorial and popular ownership that we find
at the beginning of our own history. Over and over again, at the stage in
which barbarism is beginning to pass into civilisation, the communal, in-
dividual, and signorial principles are found interwoven . . . and it seems to be
the next stage upwards in civilisation that gives its preponderance to the
lord. 1 ®
Next, to give substance to these generalisations, I propose
to quote some typical instances from Africa, Asia, and
Europe, beginning with Junod’s accouht of the South African
Bathonga.
The Bathonga system dates from before the decimation of
their cattle in recent years by disease. It belongs therefore to
an economy which was largely dependent on stockbreeding.
The plough is now used to some extent, but this is an innova-
tion. The land belongs to the chief, but only in the sense that
through him it becomes available to all who need it. Each
village headman receives from him an extensive grant of land,
the best part of which he apportions among the households
under his jurisdiction. These holdings are hereditary but in-
alienable. Land cannot be bought or sold. Similarly, when a
newcomer wishes to settle in the district, the mere act of sub-
mission to the chief entitles him to as much land as he wants,
which he proceeds to clear and cultivate. It is in the headman’s
interest to encourage him, because he enhances the value of
the land, thereby increasing the wealth and man-power of the
district, and besides he is in the habit of rendering certain
labour services . 17 Such a system presupposes a land surplus.
There is plenty of room for all comers and for shifting cultiva-
tion in each district. The Bathonga were just approaching the
economic limit of expansion in these conditions when the
British poll-tax intervened, forcing their menfolk into the
mines.
N Turning to India, we find a different and very varied set of
conditions, some of which approximate to those that have
been postulated for primitive Indo-European culture. In the
most fertile areas the soil is di&icult to clear and requires irri-
gation . 18 These factors are unfavourable to shifting tillage.
18 Hobhouse 253.
77 Junod LSAT 2. 6-7, cf. Krige 176-7, Smith and Dale 1. 387.
is Baden-PoweU 51, 66. .
304 STUDIES IN ANCIENT GREEK SOCIETY VIII
One widespread form, the raiyatwari type of village commune,
is thus described by Baden-Powell:
In the countries marked by the prevalence of villages of this type we arc
almost always able to note evidences of a tribal state of society. . . . There
were clan divisions of territory containing a number of villages, each under
its own headman or chief. . . . Each village group contains a number of
household or family holdings. ... As the headman or chief was always an
important personage, it was doubtless by his influence that the. site for
clearing and settlement was selected. . . . We find that in later times the
headman regulated subsequent extensions of cultivation and disposed of
disputes about the occupation of fresh lands. When a Raja was (perhaps in
still later days) established, it was always understood that there was no ap-
propriation of waste land without his permission, although in practice it
was often tacidy allowed and indeed freely encouraged; for the early state
authorities were only too glad to see more land cultivated, because the king’s
revenue share of the produce, which was from very early times his chief
resource, was thereby increased. ... As to the residence of the landholders, a
central village site is usually established within the group of arable lands.
In this the headman had a residence larger and better built than the others. . . .
Instances have occurred where the headman made his house a veritable fort
of refuge against marauders. . . . His office was remunerated by an important
holding of land, often the best in the village. . . . Besides this he had various
privileges and precedence rights. 19
The writer goes on to describe the status of artisans and the
conditions of tenure:
Resident craftsmen and menials are not paid by the job but are employed
by the village on a fixed remuneration, sometimes a bit of rent-free (and per-
haps revenue-free) land, sometimes by small payments at harvest, as well as
by customary allowances of so many sheaves of com. . . . The individual
holding now passes on the death of the holder to the descendants jointly,
under the Hindu law, and they divide it as far as circumstances permit. . . .
The headman alone is, or was, responsible for such village expenditure as
entertaining guests, celebrating festivals, and the like. 29
The mode of dividing the holdings may be illustrated
from villages of this type in S.W. Bengal. First, there are
. 19 Baden-Powell 9-1 5, cf. Russell 1 . 43 -4: 'The patel or village headman, on
whom proprietary right was conferred by die British Government, certainly
did not possess it previously; he was simply the spokesman and representa-
tive of the village community.' On the imposition of private ownership as. a
matter of policy by the British in India see Dutt 209-1 5.
29 lb. 16-9.
VIII
THE LAND
3°5
special allotments for privileged persons: one for the chief
of the district, another for the headman, and a third for the
priest. The remainder of the arable was divided into household
estates adjusted to their needs and periodically redistributed. 4 * * * * * * * * * * * * * * * * 21
Originally the Raja’s income was simply the produce of his
special estates ( majhhas ), tilled for him by labourers who were
granted rent-free holdings in each village, but in course of time
this was supplemented by. a general levy on the produce of the
village holdings. 22
The practice of periodical redistribution was designed to
maintain so far as possible the real equality of the holdings in
relation to the changing needs of the families. It was effected
by lot, and in some cases the procedure was very elaborate, as
may be seen in the following account from Peshawar:
The areas were taken by drawing lots. ... If the land to be allotted was
variable in quality, the dan authorities would arrange a number of cirdes or
series, consisting of good, middling, and indifferent soils, or distinguished
in some other way. Then the groups of sharers would have to take their
lands pardy out of each series. . . . But in any case, in spite of the soil clas-
sification, inequality in the holdings was not altogether excluded, and so a
system of periodical exchange or redistribution was long followed. 28
4. The English Village Community
It was the great achievement of Henry Maine to demon-
strate the affinity underlying the village communities of
Europe and Asia. A study of the pre-feudal forms of European
land-tenure enables us to draw conclusions of great value for
ancient Greece. They must of course be used with discretion,
but again and again they enable us to make sense of data which
being fragmentary are in themselves unintelligible. The
promise of this approach was recognised with characteristic
acumen by Ridgeway as long ago as 1885, when he published
21 lb. 179-80, cf. 132, 324-5, Dange 35-8. 82 lb. 181.
28 lb. 253-5, 262, 324-5. Periodical redistribution survives in parts
of the Middle East. See Warriner 1 8, 66-7, and cf. 1 9: ‘In Palestine, Trans-
jordan, and Syria still another form of semi-collective ownership exists. . . .
When the tribe settled originally, the arable land of each village was allotted
between members equally, each member receiving a piece of land in different
zones of the village; and to maintain equality between the members the
land was reallocated at intervals.'
U
306 STUDIES IN ANCIENT GREEK SOCIETY VIII
a remarkable article on the Homeric land system. Among
classical scholars it. aroused very little interest and has never
been followed up. The only further advance that has been made
in this direction is due to H. E. Seebohm, whose study of
primitive land-tenure in general gave him an immediate insight
into the Homeric problem. 24 Before availing ourselves of their
work we must prepare the ground, as they did, by studying the
land system that prevailed in our own country down to the
sixteenth century, with numerous later survivals, some of
which are not quite obliterated even to-day. In this as in other
matters wisdom begins at home.
The typical English village was surrounded by a number of
open fields or ‘shots’, each of which was divided into so many
strips belonging to different holdings. The fields were fenced
while die crop was growing, but after the harvest they were
thrown open to pasture. The meadow-land too was divided into
strips, which were distributed annually by lot among the
holders of the arable. The waste land was undivided, its use
being regulated by the community. The homesteads were
managed severally, though in early times even these were sub-
ject in some cases to reallotment. 26 In the west of England,
Wales, Scotland, and Ireland there was a different system,
known as the run-rig, in which all the I and, arable and meadow
alike, was subject to annual redistribution. In this respect die
run-rig system was the more archaic of the two. 20 It rested
direcdy on the principle that
the soil was not allotted once for all to individuals but remained in the
ownership of the tribal community, while its use for agricultural purposes
was apportioned according to certain rules among the component house-
holds, strips for cultivation being assigned by lot. 27
The lengdi of the strip varied according to the lie of the land
and the nature of the soil, but it was fixed conventionally at 40
rods, that is, 1 ‘furrow-long’ or furlong, which was as far as the
plough could be driven conveniently without a halt. The
breadth was determined originally by the number of furrows
24 Ridgeway HLS, MW, H. E. Seebohm SGTS.
26 F. Seebohm EVC 105-17, VinogradofF GM 165-6, 173.
20 EVC 438-41.
27 GM 18.
VIII
THE LAND
307
of the given length that could be ploughed in a given time —
a day or half a day. Hence the French joumel and the German
Morgen, which mean both ‘strip’ and ‘acre’. 28 The English acre
is of 'the same origin. If its length is fixed at 1 furlong, its
breadth will be 4 rods, and that was the conventional breadth
of the strip. 29
The standard unit for reckoning the size of a holding was the
hide. Its value varied in different districts, but it was com-
monly reckoned at 120 acres. 80 In Anglo-Saxon times the
holding was inalienable. 31 It was inherited by the sons, who
either worked it jointly or divided it into equal shares. This
is the rule of gavelkind, which survived in Kent. 82 It was not a
compact unit. Its component strips were scattered about in
the several shots, so that every holder had a share in the dif-
ferent qualities of soil. 33
The hide is defined by Bede as a holding sufficient for the
needs of an average family — terra unins families. As Bloch has
remarked, he was using this word in the Latin sense:
Bede’s words give us in all probability the key to the institution in its
primitive form. But we are not to think of the little matrimonial family of
our later ages. Ill-informed as we are about the history of blood relationships
in the dawn of our civilisation, there is every reason to think that the
group whose original shell was the manse was a patriarchal family of several
generations and several collateral households living around a common hearth. 34
28 EVC 124-5.
29 EVC 2. Seebohm’s identification of the strip with the acre has been
contested by Orwin 43 on the ground that its size varied; but this is
equally true of other land measures, such as the bovate, carucate and
virgate. Orwin identifies the strip with tire 'land', which is produced auto-
matically by a plough fitted with a mouldboard, such as is still com-
monly used in this country. Even if we accept this view, the dimensions of
the 'land’ remain to be explained; and it is open to the objection that, while
the strip system is found in many parts of Europe and Asia, the use of
the mouldboard seems to have been confined to N.W. Europe. The ancient
Greek plough had no mouldboard: see figs. 46, 47.
30 EVC 49-57, GM 141-4.
31 1 refer to peasant tenements, not to large estates. The latter, or rather
the rights over them, were fully alienable.
32 H. E. Seebohm 95, cf. Baden-Powell 417.
33 VinogradofFGM 175-7.
34 Bloch RDC 268. He equates the manse with the hide.
308 STUDIES IN ANCIENT GREEK SOCIETY VIII
These joint households take us back to the groups of kindred,
real or fictitious or partly both, in which the Saxons were
organised when they first landed on our shores. 86 Their names
are still enshrined in our familiar Tootings, Wokings, Eppings
and Hoppings, all based on the patronymic - ingas , 8 ® which
implies that the village had been founded by a clan or a group
modelled on the clan. 37
5. Greek Husbandry
Greece is a country of winter rains and summer droughts.
The annual precipitation rises sharply from low altitudes to
high and from south to north. In some districts there is an excess
of rainfall, which washes away the humus; in others a de-
ficiency,- which can only be made good by irrigation. We know
HG. 46. Ploughing: Attic vase
that irrigation was practised in the prehistoric period, but not
on a large scale, owing to the nature of the country, and no
important advances are recorded in historical times. 88
The holding was divided into two portions, which were
sown in alternate years. 30 There is no evidence of crop rotation
38 Chadwick OEP 303.
a® Vinogradoff GM 140, cf. EVC 346-7.
37 The evidence of the last two sections shows that Tac. G. 26 should
be translated as follows: ‘Each community occupies in turn a tract of
land proportionate to the number of its cultivators. The land is then dis-
tributed according to social status. The spacious plains make distribution
easy. The fields arc shifted annually, and there is still a surplus of land. They
do not even trouble to exploit the fertility and extent of the soil by fruit-
growing, enclosing meadows, or irrigation.’ This rendering, which implies
periodical migration in place of fallowing, is the only one that conforms to
the general probabilities of the 'case without imposing any strain on the
Latin, cf. Seebohm EVC 343-4.
38 On ancient Greek agriculture see A. S. Dorigny in Darcmbcrg-
Saglio 4. 902-10, H. Michcll 38-88, Michell gives an admirable account of
husbandry, but says nothing about land-tenure.
30 //• 18. 541, Pi. N. 6. 9-1 1, Suid. fnl kcA&ptj dpoOv.
VIII .
THE LAND
309
before the fourth century, and since fallowing is not enough
to restore the soil, it was supplemented by digging, burning, and
manuring. *0 Digging is good for vines on clay soils, but less
effective for cereals. Burning is merely a palliative. Manuring
is mentioned in the Homeric poems but not in Hesiod . 41 The
easiest way of applying manure is to turn the cattle on to the
fallow, but this method is recorded only once, in the third
century, and in several extant leases it is prohibited . 42 The
reason why it was not generally adopted is probably that the
lowland pastures, adjacent to the arable, are usually of very
poor quality, and the prohibition suggests that the fields were
not well enough enclosed to prevent die cattle from straying.
When the fallow land was brought back into cultivation, it
was ploughed at least three times. The first ploughing was done
in spring with the composite plough ( pekton arotron) drawn by a
pair of oxen . 43 The second took place after harvest, and was
done crosswise on the first . 44 On this occasion the simple
plough (autogyon arotron ) was used, preferably with mules,
which are quicker than oxen and drive a straighter furrow . 48
The third was done in October, just before the sowing . 48
The staple cereals were barley and wheat. Barley was the
easier and older, and remained the staple food of slaves . 47 It
48 X. Oic. 16. 14-5, 18. 2. 41 Od. 17. 299. 42 H. Michell 54.
43 Hes. Op. 43Z-3, 460, Oi. 1 3. 32. 44 Hes. Op. 462, Plin. NH. 18. 128.
48 II. 10. 351-3, Od. 8. 124, Hes. Op. 46.
48 In Attica, after the feast of Proerosia, held on the 5th of Pyanepsion,
he sowing month: Plu. M. 378c.
47 Semple 342-3, G. Thomson AO 2. 109-10.
310 STUDIES IN ANCIENT GREEK SOCIETY VIII
was sown in October, as soon as the rain set in. Wheat required
greater care, because the crop was liable to be ruined at the
outset by too much or too little rain, and accordingly it was
sown at discretionary intervals throughout the late autumn . 48
•The harvest was gathered in May or June according to latitude
FIG. 48. Olive harvest: Attic vase
and altitude. Threshing was not done with a flail but by
cattle treading out the grain on a cobbled floor. After being
tossed and winnowed in the cradle (liknon) the grain was
heaped in baskets and thrown against the wind, which blew
away the chaff.
The backwardness of cereals was due to the nature of the
soil, which is much more favourable to horticulture, especially
figs, vines, and olives. Figs were used largely for feeding
slaves. Olives need very little attention, but since they take
several years to mature, growers were exposed to heavy loss
from marauding raids and wars . 48 Still, olive and vine were
very profitable, as they arc to-day. The stable exports of de-
mocratic Athens, as of Minoan Knossos, were oil and wine.
Deficiency of home-grown cereals was made good by maritime
48 X.Oec. 17. 4, Thphr. HP. 8. 6. 1.
40 Semple 394, 434, Hcitland 104.
VIII
THE LAND
trade, and it is significant that the best wheat-raising areas —
Thessaly, Elis, Laconia — were for a long time politically
backward.
In most parts of the country the lowland pastures are suit-
able only for sheep, goats, and swine. Large cattle graze
£
•v*
HG. 49# County danu: Ionian va>e
throughout the summer in the mountains, and draught cattle
have to be stall-fed all through the year. Owing to the scarcity
of good grazing land cow's milk is of poor quality, and cheese
is made principally from sheep and goats. The chief sheep-
breeding areas were Thessaly, Boeotia, the Corinthian Isthmus,
and the hill country behind the Anatolian coast.
6. Modern Greek Land-tenure
The mode of husbandry which has just been outlined sur-
vives with very little modification among the Greek peasantry
312 STUDIES IN ANCIENT GREEK SOCIETY VIII
to-day. For this reason it will not be out of place to give some
attention to modern Greek land-tenure, especially to the older
forms that linger on in the more backward districts. 50
A large portion of the peasantry is now in America, driven
from home by poverty, as they were in Solon’s day. Of those
that remain many have as little as two acres to live on. In
many districts the holdings are intersected and discontinuous,
which shows that they had once been organised in strips.
The peasants live together in villages, not on separate estates.
In cases of intestacy the holding is divided equally among the
children or other next-of-kin, and the testator may not dispose
of more than the amount which after equal division would be
due to each of the direct heirs. Division among the heirs is
optional. They often decide to hold the estate jointly. This
system of joint family holdings has almost disappeared to-day,
but it was still flourishing in the last century, and we possess
a valuable account of it in Ansted’s monograph on the Ionian
Islands (1863). On the father’s death the sons and daughters
inherited equal shares in the estate, but as a rule they, did not
divide it. If they were young, they continued to live together
till they were taken away by marriage or other employment.
When the sisters married, they received a dowry equivalent to
their share of the inheritance. Some of the brothers might go
away and earn a livelihood from other sources, but they con-
tinued to pay the whole of their incomes, from whatever
source they were derived, into the family fund based on the
paternal estate. It often happened that one of them remained
at home in charge of the farm, while a second set up as factor
for the estate in the nearest town. Others might become school-
teachers or lawyers. But their incomes were united, and a close
account was kept of all transactions. When one of them died, his
share passed to his children, and, when his daughters grew up
and married, they were dowered with the share due to them
out of the joint fund without regard to their father's income. 51
5< > This subject needs to be studied in connection with Byzantine land-
tenure, on which see Ashburner FL, especially 32. 70.
51 Anstcd 1 99-201. It is not likely, of course, .that this type of house-
hold is. directly descended from the o!koj, though it is none the Jess
illuminating for that; it is probably related to the Yugoslav zadruga: Lodge
92-1 II.
VIII THE LAND 3*3
This is the system that Ansted found in Santa Mavra
(Lcukas) and to a lesser extent in Kcphallcnia and Zantc
(Zakyndios). It is remarkably like the ancient Athenian oikos
(pp. 109-12). The only important difference is that in antiquity
joint ownership terminated at the fourth generation.
7. The Open-field system in Ancient Greece
The joint family of modern Greece is only an isolated pre-
capitalist survival, but die ancient oihos was an integral unit
in the social life of die period. The city-state was a community
of oikoi. The family estate was owned by right of descent from
one of die founders of die city, and carried with it die rights
of citizenship. In commercialised cidcs like Adicns these ancient
tenures had for the most part disappeared, but at Sparta the
original estates were never forgotten, 62 and dicy must have been
remembered in many of the colonics overseas. Even at Athens,
when citizenship was granted to a foreigner, it was die practice
to enrol him in a particular tribe, phratry, and deme,* 3 and
sometimes to endow him widi a house and land. 64 Only in diat
way did he become a full member of the community. In the
Adienian law-courts we hear of persons laying claim to an
estate on the plea that an ancestor lies buried there, 66 but
■ never of disputes about land turning on evidence of sale or
purchase:
The line of argument always leads to the proof of near kinship, by blood
or adoption, to the previous owner, and the right of inheritance seems taken
for granted as following incontrovcrtibly the establishment of the required
relationship.®*!
Of course this docs not mean that estates were never bought or
sold, but that even at Athens, under a monetary economy,
deeds of transfer were not formally recognised as overriding
62 Held. Pont. RP. 2. 7.
63 SIG. 162, 175, 40, 226. 16, 310. 21, 312. 30, 353. 5, 531. 30, 543.
M Lolling IH 60, cf. ZG.2 53, D. 18. 91 etc.
„ 55 - _ I 3 _ 4 > c f. Arist. Ath. 55. 3. Conversely, the possession of
Zeus Hcrkeios, i.e. a house and land, was proof of citizenship: Poll,
o. 05, Harp. 'Epraios Ztvs, cf. Nilsson NIS 290.
86 H. E. Sccbohm 83.
314 STUDIES IN ANCIENT GREEK SOCIETY VIII
the claims of kinship. In other cities the alienation of the
original estates was actually illegal. 87
The city-state had thus arisen as a union of joint families,
each of which possessed in perpetuity a holding of land in-
herited from one of the founders. The holding had been
created at the same time as the family. It may have been subse-
quently divided, but then so was the family that owned it.
The family was bound to*the soil on which it lived. That
being so, it is incumbent on us to determine so far as we can
the manner in which the holdings were distributed.
We have already made some progress in this direction. It has
been shown that the Attic demes began as clan settlements of the
same type as the Anglo-Saxon ‘mgs’ and ‘hams’ (pp. 1 12-3); and
that the rule of succession to the oikos corresponds to the Anglo-
Saxon law of gavelkind. We must also, of course, remember that
the Old English land system was by no means peculiar to this
country. It occurs in analogous forms in all parts of Europe,
India, China, Central and South America. 88 It is in fact charac-
teristic of the primitive village community; 88 and if we are to
approach Greek land-tenure with anything in our minds at all,
this is the institution we should keep before us and not the
boundary squabbles of twentieth-century gentleman-farmers.
At Athens under the democracy it was a regular policy to
relieve unemployment and at the same time to secure strate-
gical points by settling poor citizens overseas on conquered
territory. 80 The land selected was divided equally into as many
holdings as there were citizens enlisted under the scheme, and
the holdings were then distributed by lot. The settlers were
required to reside within the territory, but as a rule they did
not work the land themselves. That was done by the native
proprietors, who were left in occupation subject to the pay-
ment of an annual rent. The best-known of these klerouchlai or
‘lot-holdings’, as they were called, is the plantation of Lesbos
in 427-426 B.c. In the previous year the people of this island,
with the exception of Methymna, had revolted against Athenian
87 Arise. Pol. 13 x9a. 9, Hdd. Pont. RP. 2. 7.
88 F. Sccbohm EVC 186-206, 214-62, 336-88, Skene 3. 139, Kovalevsky
162-70, Wittfogel 348-409, Bancroft 2. 226, Thompson 49.
88 H. E. Sccbohm 88. 88 Grundy 177-8, 201.
VIII
THE LAND
315
rule. The leaders were executed, and the island was saddled
with a plantation, which Thucydides describes in the following
words:
They divided the land, except Methymna, into 3000 allotments, of which
300 were set apart and consecrated to the gods and the remainder settled
with lot-holders from Athens. The Lesbians continued to till the soil,
subject to an annual rent of 2 mnat on each allotment.® 1
These allotments were equal in size. That was inherent in the
nature of the scheme, and is proved by the uniformity of the
rent. The natives remained in occupation. How then was the
land divided among the Athenian lot-holders? If, as has been
generally assumed, the island had previously been cultivated
in separate estates, enclosed and consolidated like modern
capitalist farms, their size would have varied indefinitely. It
would consequently have been impossible to divide them into
equal lots without drastically reorganising the native tenures.
But according to Thucydides that was not done. Another
course would have been to graduate the rent according to
the size of the farm and divide the income among the lot-
holders. But again that was not done. The only conditions that
meet the requirements of the case are those of the primitive
village community. With a unit of division ready to hand in
the strip, it would have been possible to combine or divide
holdings of different sizes into equal lots without disturbing
the existing tenures. Each village area would be assessed for so
many lots, and the work of sharing out the liability could be
left to the villagers themselves, as in India.
This conclusion is so far-reaching in its implications that
it would be unwise to insist on it beyond inviting the attention
of historians to a problem that has apparently escaped them;
but there are one or two considerations arising from the present
issue that may conveniently be mentioned here. There is no
reason to suppose that the procedure at Lesbos was abnprmal.
We have another instance in the plantation of Chalkis (Euboia)
eighty years earlier (506 B.c .). 62 In this case the number of
lots was 4000. At the time in question the landed nobility of
Euboia were still in power, and it was these landowners, not the
cultivators, that the Athenians displaced. They had been more
61 Th. 3. 50, cf. SIG. 1. 76. 62 Hdt. 5. 77* 2.
3 16 STUDIES IN ANCIENT GREEK SOCIETY VIII
successful in maintaining themselves than the Attic land-
owners because in Euboia the struggle for the land had been
relieved by colonial expansion. This shows that, granted the
possibility of colonisation, there is no inherent difficulty in
supposing a primitive land system to have coexisted with rapid
commercial development. At Lesbos, it is true, the democratic
movement began much earlier, but on the other hand it was
arrested in the sixth century by the Persian conquest . 63
The farmers of Lesbos were reduced by the plantation to the
status of a rent-paying peasantry. A similar problem arises in
regard to odier Greek states, which had been founded on a
tributary peasantry from the start. When the Dorians con-
quered Sparta, they did not dispossess the natives of the
country, who continued to live in their ancient villages . 64
Nevertheless they divided the land among themselves into in-
alienable family estates, cultivated for them by the natives,
who were forced to surrender fifty per cent of their produce . 66
And these estates were equal, that is to say, their size was
adjusted at the foundation to the needs of the proprietors, who
had to provide from them their contributions to the common
meals . 66 It is clear, however, that, whatever may have been die
case in Lesbos in the fifth century, there cannot have been any
extensive appropriation of the land at Sparta in the eleventh.
Here again, therefore, the fact that the natives remained in
occupation is a sign that the new holdings were distributed on
the basis of the strip system. Stabilised at this early date by the _
act of conquest, the Spartan aristocracy was exceptionally suc-
cessful in resisting change. A further proof of its primitive
character is furnished by the configuration of the city itself.
Even in the time of Thucydides it was not properly speaking a
city at all, but a group of adjacent villages . 67 It is in keeping
widi the general probabilities of the case that this rudimentary
degree of urbanisation was combined with the survival in
tributary form of the primidvc village commune.
63 We have several inscriptions from Lesbos (Roman period) giving
lists of farms with the acreage under com, olive, vine, and grass, and the
size of the farms varies indefinitely: IG. 12. 2. 33-7.
64 Liv. 34. 27, cf. p. 393.
66 Held. Pont. KP. 2. 7. Arist. Pol. 1270a, Tyrt. 5.
66 Plu. Lye. 8, Plb. 6. 45. 3. 67 Th. 1. 10. 2.
VIII THE LAND 317
The origin of the science of geometry is explained by
Herodotus:
King Sesostris divided the land of Egypt among the people so that each
received a square allotment of equal size, and from these holdings the king
drew his revenue by an annual impost. If part of a holding was swept away
by the Nile floods, the proprietor informed the king, who would then send
his overseers to measure the loss and the tax was reduced accordingly. This,
in my opinion, was the source from which the Greeks acquired die art of
measuring land (garnffWa).® 8
The Greeks may not have been indebted to Egypt so directly
as Herodotus supposes, but his main point is proved by the
word itself. The starting-point of geometry was the need to
divide the land.
We have seen how the land measures of western Europe —
the acre, joumel, and Morgen — were based on the dimensions
of the strip. There is an analogous term in Greek, which will
help us to reconstruct the dimensions of the Greek strip.
The word gyes, used in Homer as a land measure , 69 means
properly ‘plough-tree’. It was also applied to the primitive
type of plough, consisting simply of a forked bough, such as
may still be seen in parts of the country. We may infer, with
Ridgeway, that as a measure the gyes denoted originally a
plough-acre, that is, the amount of land that could be ploughed
in a given period. The period was probably a day, because one
of the Homeric words for ‘evening’ is loulyt 6 s, the time for
‘unyoking the oxen’.’®
The ancient commentators inform us that the gyes was
equivalent to one pUtbroit . 71 This was a long measure equal to
100 feet . 72 So the gyes was a plough-acre which measured 100
feet along one of its sides. Which side was this? There is
another Homeric land measure, the oAron. We read of an
‘oAron of oxen’ and an * oAron of mules', the latter being the
longer . 72 The word is probably a heteroclite form of oAros
‘boundary’, which again is connected with oureAs ‘mule’ and
Latin urvutn ‘plough-tail ’. 74 The oAron of mules is explained by
the commentators as ‘the amount of land that a mule can
68 Hdt. 2. 109, cf. x. 66. 2. 69 11 . 9. 579, Oi. 7. 113, 18. 374.
7 ® 11 16. 779, 01 . 9. 58. II. 9. 579 sch. 72 II. 2i. 407, 01. 11. 577.
78 11 . 10. 351-3, Oi. 8. 124-5. 74 Boisacq f.vv.
3 18 STUDIES IN ANCIENT GREEK SOCIETY VIII
plough at one effort, that is, a pUthrori.™ From this we see that
the g yes and the ouron are identical. They are the plough-acre,
which measured ioo feet across the furrows, the length of the
furrow being given,
For the length of the furrow we have only one clue. Remem-
bering that in all countries units of long measure are commonly
based on tillage, we note that in Greek 6 plethra make I
stadion, that is, 600 feet. The stadion was the standard long
measure, which yielded the word stadium or racecourse. The
racecourses at Olympia and elsewhere all measured 600 feet in
length. 70 Now, in Argive Doric the form ''of the word is not
stadion but spddion. These are not phonetic variants. They are
different words. And both are apt designations for the length of
the furrow, for sta means ‘stand' and spa means ‘puli’, referring .
to the distance the plough is drawn by the ox or mule before it
is halted and turned. The Greek stadion is therefore a unit of the
same origin as the English furlong. This hypothesis is con-
firmed when we find that the breadth of the Greek racecourse
was generally about 100 feet. 7 7 The original racecourse was a strip.
And now after this long but not unprofitable digression let
us return to the passage in Homer with which we began:
Like two men with measures in their hands quarrelling over boundaries in a
common ploughland, contending for equal shares in a small space of ground;
so the two sides were parted by the battlements, over which the warriors
slashed at one another's shields as they fought. 70
As the breadth of the strip, the_ oviron was the distance from
balk to balk. The Greek balk was a row of stones ( oiiroi ) such as
are still to be seen in Palestine, as they were when the children
of Israel were warned not to remove their neighbour’s landmark. 7 8
The connection between ouron, the width of the strip, and /
oilros, the row of stones which separated one strip from another,
76 11 . 10. 351 sch. AV.
70 Pauly- Wissowa 2. 5. 1969.
77 Pauly-Wissowa l.c. It seems probable that the strip was the unit
underlying the 3vy6v, a land-measure attested for Amorgos in the fourth
century (SIG. 963). In Byzantine Greek the jeuy oy was the amount of land
that could be ploughed by a pair of oxen in a day: Cod. Just. 1 o. 27. 2. The
modern orptyiia seems to be based on a Turkish unit of the same nature.
78 //. 12. 421-5. ,
70 II. 21. 403-5, Deut. 19. 14.
VIII
THE LAND
319
is explained. And this row of stones answers to the embattled
parapet over which the Greeks and Trojans are locked in com-
bat. The comparison is apt. The two men in the simile are
marking out the shares that have been allotted them in one of
the open fields. The field is not a large one — perhaps it has
already been encroached on by private enclosures; and so each
of them is bent on getting his foil share. But they are not the
owners. They are dividing it merely for use. Perhaps some day
it will be redistributed. And so it is described as ‘common’ in
the not uncommon sense of being owned in common by the
village commune to which these survivors of primitive com-
munism belong.
8. Redistribution of the Land
It is quite possible that, at the time when the Iliad and
Odyssey were put into their final shape, the custom of periodical
redistribution was becoming obsolete; but Homer does not set
out to tell us everything, and before drawing any conclusions
we shall do well to review the evidence.
Early in the sixth centuiy, when the Attic countryside was
seething with unrest, Solon introduced a number of agrarian
reforms, which tided over the crisis but failed to satisfy the
peasantry, because they had demanded a 'redistribution of the
land’. 80 A procedure of this kind was actually carried out at
Kyrene, a Greek colony on the coast of Libya. Some time in the
sixth century new settlers from the mother-country were in-
cited to participate in a ‘redistribution of the land’. The whole
people, including the newcomers, was divided into three
tribes, and on this basis, after special estates had been set aside
for the king in his capacity as chief priest, the soil was re-
allotted. 81
The demand of the Attic peasants was not, as usually in-
terpreted, a revolutionary one — a subversive challenge to the
sacred rights of private property. It was counter-revolutionary —
a protest against the appropriation of the land, which was
so Arist. Atb. 11. 2. A redivision of the soil was demanded at Leontinoi
in the fifth century B.c.: Th. 5. 4. 2 .
81 Hdc. 4. 159-61, cf. Th. 8. 21.
320 STUDIES IN ANCIENT GREEK SOCIETY VIII
violating the sanctity of the old communal rights. It was an
appeal to the past, not the future. As we see from the case of
Kyrene, the principle of redistribution was still alive. It only
remains to show that it had once been periodical.
The Greeks were well acquainted with this practice. Strabo
says that the Dalmatians reallotted the soil every eight years, 82
and among the Vaccsi of Spain, according to Diodoros, it
was done annually:
The Vaccaei divide the land every year, each receiving a portion of the
fruits, which are common property. Appropriation is punished by death. 83
•
Early in the sixth century a band of Dorians from Rhodes and
Knidos set sail for Sicily. Their intention was to plant a colony
at Lilybaion, but they were driven off by Phoenicians. Then
they sailed to the Liparai Islands, where they joined forces
with the natives. The rest of the story may be told in the
words of Diodoros:
Being well received at Lipara, the settlers agreed to share the land with the
inhabitants. ... In course of time, owing to the depredations of Etruscan
pirates, they built a fleet and divided their occupations. Some of them con-
tinued the collective tillage of the soil, while the remainder organised them-
selves for defence against the pirates. They held their property in common
and ate at common meals. After leading this communal life for some time,
they divided Lipara itself, where the town was, but continued to cultivate
the other islands collectively. Eventually they divided all the islands for
periods of twenty years, reallotting the land at the end of each period. At sea
they won a number of victories over the Etruscans, and from die spoils sent
many memorable tithes to Delphi. 84
Diodoros has given us more than we asked for. Besides recording
a system of periodical redistribution in a Greek city-state,
he takes us back to a still earlier stage in which the land had
been owned and cultivated by the village communes without
even a temporary division.
The reader will ask how modern historians, especially
Toutain, whose condemnation of the slave-minded Esmein is
still fresh in our minds, have succeeded in interpreting this
passage otherwise. Toutain takes the liberty of not mentioning
it at all. So does the writer in the Cambridge Ancient History
who assures us that the Greeks had long outlived the stage
33 Str. 315. 83 D.S. 5. 34, cf. Nic. Dam. iz6. 84 D.S. 5. g.
VIII
THE LAND
321
which Diodoros describes. Presumably they were both satis-
fied that the question had been settled once and for all by
Guiraud in his La propriltl foncilre en Grice, still the standard
work on the subject, to whose Vigour and clarity’ Toutain
repeatedly appeals. Let us sample this Vigour and clarity’:
We have no reason to question the veracity of Diodoros in - regard to the
existence of agrarian collectivism in the Liparai Islands. The only doubtful ~
point in his account is the motive assigned for the adoption of the system.
Thfodore Reinach has recently shown from a passage in Livy that, like the
Etruscans, the Liparai islanders were pirates. That being so, it is not dif-
ficult to perceive that so far from being a relic of the past their communism
was an artificial regime created for a specific purpose. No political or social
principle is here involved. These islanders simply adopted the institutions
most appropriate to a band of brigands. . . . Furthermore, a breach was soon
forced in the system by the love of private property, so potent in mankind.
In the course of the fifth century at latest they began to divide the main
island, which was doubtless the only one fortified and inhabited. The others
remained undivided. 86
Before getting down to 'political and social principles’ let us
make sure of the facts. Diodoros assigns no motive for the
system. It is Guiraud who has done that. There is nothing to
show at what date the main island was divided. Diodoros
evidently believed that some at least of the other islands were
inhabited, and Strabo, our only other authority, agrees with
him. He names eight islands in all and describes only two of
them as uninhabited . 80 Guiraud's interpretation of the passage
is thus contradicted by the passage itself. It is true that Livy
describes the islanders as pirates, and no doubt they were.
Guiraud does not quote his actual words, but he who runs may
read. Mos erat civitatis velut publico latrocinio partam praedam
dividend ‘ Their custom was to divide the spoils, which were
acquired by a sort of collective brigandage.’ These Greek com-
munists were at least consistent. After piously dedicating a
tithe to the gods, they shared the rest among themselves. All
property, personal and real, acquired and inherited, was
collective. The picture of primitive communism is complete.
Having rescued the facts, we may enquire into the principles.
Being no better than robbers abroad, these islanders had
naturally failed to develop at home that respect for private
as Guiraud 1 3-4. 80 Str. 275-7. 87 Liv. 5. 28. 3.
W
322 STUDIES IN ANCIENT GREEK SOCIETY VIII
property which Guiraud, like Sir John Sinclair, regarded as the
hallmark of civilisation. But it is rash to assume that their
institutions involve no ‘political or social principle*. If these
unprincipled islanders were pirates, so, whenever they had
the chance, were the Etruscans, Carthaginians, Phoenicians,
Carians , 88 and all the seafaring peoples of antiquity, including
the Greeks themselves . 89 The Achaean heroes of the Iliad and
Odyssey were pirates, and proud of it , 60 and, as we shall see
shortly, they shared out their ill-gotten gains in the same way . 91
The civilised Greeks, whose political and social principles
have been held up by Guiraud and others as a model for man-
kind, saw nothing in privateering that was incompatible with
the honour of a gentleman, and in numerous treaties they
made express provision for the exercise of piratical rights . 92
After all, there is no difference in principle between sea-
raiding and land-raiding. All pirates, all raiders, all conquerors,
all empire-builders, no matter how fervently they bow down
before the sacred presence of private property, once they have
got it into their own hands, begin, like the men of Lipara, by
stealing it. If Guiraud had allowed this train of thought to roll
on uninterrupted, he would have found himself face to face with
the political and social principle that his own civilisation rests
on robbety. La proprifte cest le vol . 83
Having ‘interpreted otherwise’ all the evidence for common
ownership that has been drawn from this and other Greek
sources, Guiraud concludes:
One needs to have a singularly biased mind to attach the least value to
them. There is not in the whole of ancient literature a single passage which,
sanely interpreted, confirms such an assertion. 94
If sanity is freedom from bias, we are all to some extent de-
fective. It is a matter of degree. But at least there is one social
and political principle that this vigorous historian has suc-
ceeded in establishing. He has demonstrated with disarming
clarity that the love of private property in the modern bour-
8 8 D.S. 5. 9, PIb. 3. 24. 4, Th. 1. 4, 7-8, Oi. 15. 415-84.
8 » Hdt. 1. 166, 6. 17, Th. 1. 5, D. 50. 17, Lycurg. Leo. 18.
11 . 11. 625, OL 4. 81-90, 9. 40-2, 14. 229-34.
9i See p. 329. 92 Hasebrock 1 17-21.
93 Engels UFPS 127. 94 Guiraud 21-2.
VIII
THE LAND
323
geoisie is a force so potent that they cannot imagine life
without it:
Was ihr niche fa sst, das fehlt euch ganz und gar,
Was ihr nicht rechnet, glaubt ihr, sei niche wahr.
9. The Method of Distribution
Let us return to the klerouchfa. After the number of settlers
had been fixed, the land to be settled was divided into the same
number of lots. How the settlers were chosen we are not told,
but it must have happened sometimes that the number of
applicants was in excess of the available land, and, since
sortition is known to have been used extensively under the
democracy, it may be assumed that the applicants drew lots.
We must also suppose that the holdings were subject to the
same rules of inheritance as prevailed in the home country.
Each lot-holder became the founder of an oikos, a family estate
of the normal type, except that being new it was unencumbered
with hereditary claims and so easier to alienate. We know that
in many cases, contrary to the regulations, the lot-holders sold
out and returned home.® 6
The klerouchh differed from a colony of the ordinary kind
( apoikia ) in that its members retained all their rights as Athenian
citizens. 06 The colony was a new city-state, bound to the
metropolis by religious ties but politically independent. Apart
from this, the mode of organisation was the same. Kyrene was
colonised from Thera in the seventh century at a time when
that island was suffering from a famine. One of every pair of
brothers throughout the island was chosen by lot. 97 A similar
procedure was adopted by the ancestors of the Etruscans when
they left Lydia. In this case too there was a famine. The king
divided the people into two equal portions and cast lots
between them. 88 It was evidently a traditional practice for the
founders of a colony to be selected by lot.
In the middle of the fifth century an Athenian colony was
founded at Brea on the Thracian coast. The decree regulating
the procedure has survived. One of the provisions is that 'ten
men, one from each tribe, shall be elected as land-sharers, and
06 Grote 6. 37-8. 00 E. M. Walker in CAH 4. 161.
07 Hdt. 4. 153, cf. Parch. 5. ' 08 Hdt. 1. 94. 5.
324 STUDIES IN ANCIENT GREEK SOCIETY VIII
these shall allot the land ’. 89 The land-sharer (geonSmos) dis-
tributed the holdings, as distinct from the land-measurer
(geometres) who marked them out. From this we gather that the
holdings were distributed by lot, and that they were co-
ordinated in some way with the tribal system. The latter point
is confirmed by Plato's regulations for the foundation of his
ideal state. In this case of course the situation is imaginary, but
it is generally recognised that he modelled his procedure on
actual practice:
In the first place, the city is to be located as nearly as possible in the centre
of the territory. . . . Next, the whole area, including the city, is to be
divided into twelve portions, starting from an enclosed sanctuary of Hestia,
Zeus, and Athena, which shall be called the Acropolis. These portions are
to be equal and adjusted in extent to the quality of the soil. Altogether
5040 holdings are to be described, and each of them is to be divided into
two parts, one in the city, the other at a distance. . . . The citizens them-
selves shall then be divided into twelve groups and a full inventory made of
all their personal property so that this too may be divided as evenly as pos-
sible among the groups. Finally, each group is to set apart an estate for one
of the Twelve Gods, and it shall be called a tribe. 1 *’ 0
Plato's total of holdings is reached by the same sort of mystical
progression as Plutarch’s formula for cat and kittens (p. 214):
1 X 2 X 3 X4X 5 X 6x7=5040. But even this is founded on fact.
It is just over half of what was regarded as a suitable total. 101
The Athenian colony at Amphipolis was divided into 10,000
holdings, and the Syracusans fixed the same number for their
colony at Aitna. 102
Plato’s lot-holders were heads of families. That is clear from
other passages in the Laws. But these families are treated as
components of a larger unit, the tribe. This too is confirmed
from other sources. In early Rhodes there were three settle-
ments — Lindos, Ialysos, Kameiros. 103 They corresponded, as
00 Tod 88-90. 100 PI. Leg. 745.
101 Arist. Pol. 1267b. 3. 102 Th. 1. 100, D.S. 11. 49.
103 2 /. 2. 655-6, Pi. O. 7. 73-4, SIG. 339 n. 2. According to the Iliad
the founder was Tlepolemos from Ephyra, probably the Thessalian
Ephyra, cf. D.S. 5. 58; but in Pindar the island is settled by the three sons of
Helios, and in yet another version by Althaimenes from Crete: Apld. 3. 2.
1-2. There were in fact several successive settlements: Str. 653-4. We
have a list of Rhodian clans grouped in phratries, and some of them have the
Aeolic termination -hoi: IG. 12. 1. 695.
VIII
THE LAND
325
we learn from the Iliad, to the three tribes of the immigrants.
Further, they were assigned by lot. This is implied by Pindar’s
account of them; for in the same poem he relates how the
gods had cast lots for the division of the world. The Sun-god,
he says, happened to be absent at the time, and so was left
without a portion. The omission was rectified by assigning him
the island of Rhodes, then beneath the sea, which he had
descried rising to the surface, and this arrangement was
ratified by Lachesis, the goddess of Allotment. 10 * The allot-
ment of the newly-conquered world among the Sons of
Kronos is presented as a divine precedent for die allotment of
the newly-conquered island among the Sons of Helios, the
founders of the three tribal setdements that bore their names.
So far we have found no mention of the clan. It is not hard to
see why. In the mature city-state the tribe persisted as a
military and political unit long after the phratry had dwindled
into a purely religious union and the clan had dissolved into
families. The clan survived to some extent among the aris-
tocracy, but in general the colonies were recruited from the
lower classes — from men who wanted land. 105 And this was the
section of society in which clan ties had most completely dis-
appeared. If we want to find traces of the clan, we must go back
to the prehistoric period.
We have seen that the Athenian klcrouchla conformed in
principle to the mode of organisation that had been followed in
the great period of colonial expansion, from the eighth to the
sixth century. This was the movement that scattered the
Greeks over the whole of the Mediterranean. And now, •
arguing still further back, we can see that, since these colonies
reproduced the structure of their mother-cities, they were a
continuation of those still earlier movements that had estab-
lished the mother-cities themselves in Greece and the Aigean.
The Greeks recognised the continuity. They remembered
how, in the days before the Trojan War, Tlepolemos, the
founder of Rhodes, ‘effected an equitable partition of the
land’; how, earlier still, Makareus had ‘divided the soil' of
104 Pi. O. 7. 54-76, cf. Apld. z. 8. 4, Paus. 8. 4. 3.
iob in the Brea decree ( 1 . 41) it is stipulated that the colonists are to be
recruited from the poorest classes, cf. PI. Leg. 735-6, Iso. 4. i8z.
326 STUDIES IN ANCIENT GREEK SOCIETY VIII
Lesbos; how Kydrolaos had ‘settled in Samos and divided the
land into allotments'; and how Tenedos, the eponym of that
island, had allotted the soil among his people and received for
himself a special estate (tlmenos) in which he was worshipped
after death as a hero. 106 Clearly, these prehistoric settlements
were not city-states at all but tribal federations. In them there-
fore we should expect the clan to have been more prominent.
And so it was. The Ionian colony of Teos, founded after the
Dorian invasion, was divided into pyrgoi or demes, and as late
as the fifth century several of these were still inhabited by the
clans after which they were named (p. 169). The three ter-
ritories of Rhodes were also divided into demes. One of these
was Netteia of the Nettidai, another was Hippoteia of the
Hippotadai. 1 07 This evidence confirms the conclusion drawn from
our examination of the Attic demes (pp. 112— 3). In Attica, it
is true, we know of only one clan — the Boutadai (p. 1 08) — which
continued to reside in its ancestral deme. This is because the
country had been transformed by the social upheaval of the
sixth century. But even in Attica the old ties, though they had
been severed, were not forgotten. Kimon of the Philaidai be-
longed to Lakiadai between Athens and Eleusis; but he must
have known — otherwise we should not — that his forefathers
had come from Philaidai near Brauron, where Philaios first set
foot on Attic soil (p. 121). And there was a tradition that,
when King Theseus was reorganising the country, he made a
tour of the rural districts, ‘visiting the demes and clans’. 108
The implication is that in those early days the two units
were identical. This conclusion has already been argued in
Chapter IV (pp. 112-3) and final confirmation is forthcoming
from the word itself. In Homer demos denotes both a tract of
cultivated land and the people inhabiting it. 109 It is properly
a ‘division’, being cognate with dasmos, which was the word
100 D.S. 5. 59, 81-3.
107 SIG. 932. 24, 33, 1 18. 5, 695. 21.
108 Plu. Thu. 24.
100 ll. 5. 710, 20. 166 etc., cf. t&Spos (1) ‘estate' (2) ‘heirs to the estate’,
Arabic hayy 'tribe' or 'tribal territory’ (Robertson Smith KMEA 39), Anglo-
Saxon lid 'household' or ‘household estate’ (Vinogradoff GM 141), Ramsay
85-6, Skene 3. 137-7.
VIII
THE LAND
327
regularly used for the division or distribution of the soil . 110
In origin, therefore, the deme was a unit both territorial and
social — a clan settlement, like the English Woking, Tooting,
Epping, the French Aubigny, Corbigny, Pontigny, the
German Geislingen, Gottingen, Tubingen , 111 and the Hebrew
‘families’ or clans which settled in the Promised Land:
Speak unto the children of Israel and say unto them, When ye are passed
over Jordan into the land of Canaan, then ye shall drive out all the in-
habitants of the land from before you. . . . And ye shall divide the land by
lot for an inheritance among your families; and to the more ye shall give
the more inheritance, and to the fewer ye shall give the less inheritance;
every man's inheritance shall be in the place where his lot falleth; according
to the tribes of your fathers ye shall inherit. 112
And Joshua said unto the children of Israel, How long arc ye slack to go to
possess the land which the Lord God of your fathers hath given you? Give
out from among you three men from each tribe; and I will send them, and
they shall rise and go through the land, and describe it according to the in-
heritance of them; and they shall come again to me. ... Ye shall therefore
describe the land into seven parts, and bring the description hither to me,
that I may cast lots for you here before die Lord our God.ua
10. The Growth of Privilege
The Greek for a holding is kleros, a 'lot'. In poetry moira
‘share’ and Idehos ‘portion’ are used in the same sense. All these
words are Indo-European. The primary meaning of kleros was
a ‘piece of wood’, like the Irish cldr, ‘board’ or ‘beam’, showing
that chips of wood had been used for casting lots. The base
kla recurs in the Greek klddos ‘branch’ and kldo ‘break’, also in
the Gothic hlauts, which is cognate and synonymous with the
English ‘lot’. As these etymologies show, the use of the lot was
an ancient feature of Indo-European culture. It rested on the
uo Boisacq s.v. Bifros. Modern scholars, having neglected to analyse the
structure of tribal society, arc necessarily blind to this inherent connection
between clan and village. Thus, following Gardner (p. 106), F." E. Adcock
excludes the clan altogether from his remarks on the origin of the tt6Ais (‘the
clan, the ginos, which is a reflection of aristocracy, is yet in the future’); and
consequendy, faced with the fact that the tribesmen lived in villages, he can
only remark chat they did so ‘as by instinct’ (in CAH 3. 688).
111 F. Secbohm EVC 355-6 7.
112 Num. 33. 51-4.
113 Josh. 18. 3-6.
328 STUDIES IN ANCIENT GREEK SOCIETY VIII
principle that every member of the community was entitled
to an equal share in the product of its labour. 11 *
In early Greece this principle had already been limited by
the custom of reserving portions of land for the special benefit
of priests, chiefs, and kings. In the plantation of Lesbos a
FIG. 50. Casting lots: Attic vase
tithe of the holdings was ‘set aside* for the gods (p. 315). The
settlers at Brea were granted the whole of the land with the
exception of certain estates 'set aside’ for the priesthood. 115
Similar estates were ‘set aside’ at Kyrene for the king. 116 In the
Odyssey , when King Nausithoos led the Phaeacians to their new
home, he fortified a city for them, divided the ploughlands,
and built temples for the gods. 117 The Homeric poems make it
clear that, while various privileges were in the gift of the king,
the land was controlled by the people. Bellerophon was re-,
warded by the King of Lycia with royal honours, but his
estate of rich arable land was bestowed on him by the people. 118
./Eneas was warned by Achilles, whom he had come to fight,
that, even if he should win, he could not hope for honours
from Priam, who had sons of his own to provide for, nor for an
114 In Greece, the lot remained in use for elections, determining priority
of approach to the Delphic Oracle, and the appointment of clan chief s:
A. I. 32, Tocpffer AG 21, 123, Paton 137.
no Tod 88-90, cf. A E. 403-5. no Hdt. 4. 161. 3.
“ 7 01 6. 9-10. ns H. 6. 193-5.
VIII
THE LAND
329
estate from the people. 119 The elders of Aitolia, presumably the
'• clan chiefs, tried to induce Meleagros to fight for them by
offering him an estate in the most fertile part of the country. 180
These reservations are called temfnea, estates ‘cut off’ (timno) or
‘ set aside’ ( txairio ) from the remainder of the land, which was
divided among the people. The timenos is the germ of private
property emerging within the tribal system.
The same combination of communal with individ ualis tic
principles appears in the sharing of booty. The process of dis-
tribution is the same — a Aasmis effected by lot; and just as the
king is granted a special holding of land, so in the division of
spoils he receives a special ‘privilege’ (g/ras) or ‘reward’ (time)
reserved from the general allotment. 121 The disguised Odysseus
boasts of having led nine raids, from each of which he received
generous gifts over and above his share in the distribution , 188
After plundering the town of Thebe, the Achasans ‘divided
the spoils and set aside the daughter of Chryses 'for Agamem-
non'. 123 Later on, forced to restore the girl, Agamemnon de-
manded compensation, but, as Achilles reminded him, it was
too late:
How can the Achaans give you a glrasl The spoils we took have already
been divided, and it would not be right for the people to bring them to-
gether again. Let the girl go, and if ever Zeus grants to us the sack of Troy,
you shall be repaid threefold and fourfold. 184
This principle that acquired wealth was subject to popular dis-
tribution was very tenacious, and not only in remote corners like
the Liparai Islands (pp. 320—1). As late as 484 b.c. the Athenians
proposed to share out a surplus from the silver-mines among
110 II. zo. 178-86.
12 0 ll. 9. 574-80. The elders were presumably the clan chiefs: Glotz
SF 12. The tlmenos must have included- slaves to work it: Jeanmaire 75,
cf. ll. 9. 1 54-6. The conditions of tenure are doubtful; probably it went with
the chieftaincy.
121 ll. 1. 166-7, 368-9, 2. 226-8, cf. h.A. 945, E. Tr. 248, 273, cf.
Bancroft 2. 225, Baden-Powell 195, Robertson Smith KMEA 65.
128 Od. 14. 229-33. In the Aegean, as late as the end of the eighteenth
century A.D., when a ship returned from a cruise, mercantile or piratical,
the proceeds were divided into two portions, one of which went to the share-
holders of the ship, while the other was divided equally among the crew:
Melas 35.
123 ll. 1. 368-9. 124 17 . 1. 123-9.
33 ° STUDIES IN ANCIENT GREEK SOCIETY VIII
the whole citizen body. Themistokles persuaded them to
spend it on building a fleet instead . 125 The old tribal custom ■
was incompatible with the growing interests of the state.
As with booty, so with food. In early times, so Plutarch
writes, when meals were administered by Moira or Lachesis
on the principle of equality, everything had been decently
and liberally arranged; and in support of this contention he
points out that the old word for a feast meant properly a
'division '. 120 His etymology is correct: dais is cognate with
dasmSs. The moirai of meat were divided equally. When the
disguised Odysseus entered his home, the meat was being
served for the evening meal, and Telemachos insisted that the
beggar was to receive a portion 'equal to those which had been
allotted to the suitors ’. 127 In the Homeric Hymn to Hermes
the meat offering for the Twelve Gods is cut into twelve
portions, which are distributed by lot . 128
On the other hand, the chine, which was the choicest por-
tion, was reserved as a geras for the chief presiding at the meal.
When Menelaos invited his guests to sit down to table, he
handed them the chine which his servants had placed before
him . 129 The swineherd paid the same compliment to the dis-
guised Odysseus — a dramatic touch, because he gives his lord
the lordly portion without knowing who he is . 130 When
CEdipus was served by his sons with the haunch instead of the
shoulder, he cursed them . 131 At the shrine of Amphiaraos at
Oropos it was the rule that, whenever a victim was sacrificed,
the priest was entitled to the shoulder . 132 The prerogatives
enjoyed by the Spartan kings were almost entirely of this kind.
135 Hdt. 7. 144.
138 Plu. M. 644a, cf. Od. 8. 470, Hes. Th. 544, Thgn. 677-8. On the
public feasts of ancient Greece, which were a modified survival pf the
primitive communal meal, see Fustcl de Coulanges 179, Nilsson HGR
254-5, Robertson Smith RS 282.
127 Od. 20. 279-82. 138 Horn. H. 4. 128-9.
130 Od. 4. 65-6, cf. ll. 7. 321, Horn. H. 4. 122.
139 Od. 14. 433-8. 131 S. OC. 1 375 sch.
132 SIC. 1004. 30-I. In Kos special portions of meat offerings were
reserved for particular clans: Paton 88-90, cf. SIG. 271, 589, Plu. M.
294c, D.S. 5. 28. For modem examples see Krigc 55 -6, Junod x. 329,
Earthy 37, 159, Roscoc B (1923) 165, Hutton 75, Gurdon 48, Ivcns
MSE 408.
VIII
THE LAND
331
Both of them had a royal estate and a priesthood. At the
opening of a campaign they sacrificed as many sheep and goats
as they pleased and kept the chines and hides for themselves.
At all common meals, whenever present, they were presented
by the company widi two quarts of barley meal and half a pint
of wine. 133 As Thucydides remarks, the Greek kingship rested
on ‘defined prerogatives’. 134 And the conditions on which they
were enjoyed are stated in a famous passage of the Iliad:
Why have the people of Lycia conferred on us the highest honours — pride
of place and precedence in food and drink? They look on us as gods, and
they have bestowed on us a timtnos of rich ploughland. Therefore we must
be foremost in the fray, that the people may say, These kings of ours, who
feed on our fat herds and quaff our choicest wine, can fight. 136
Royal honours were bestowed by the people in recognition of
military service.
The Greek for 'prerogative' is giras ; the Greek for ‘old age’ is
gcras. The origin of the privileges accorded to the tribal elders
was discussed in Chapter I (p. 45). In a hunting economy they
took the form of exemptions from dietary taboos and prescriptive
shares of game. In the course of ages, as these elders evolved into
magicians, chiefs, priests or priestesses, kings or queens, their
rising status was continuously accommodated to the traditional
forms. The giras of meat, the giras of booty, and the giras or
timenos of land, reflect through hunting, warfare, and tillage the
emergence of social inequalities out of primitive communism.
The use of the lot was a guarantee of equality. The distri-
bution was made impartial by placing it beyond human con-
trol. And being beyond human control it was regarded as
magical — as an appeal to the Moirai or 'Portions’, the god-
desses who determined each man’s lot. In origin these figures
are simply mythical projections of the practice of casting lots.
How then did these equalitarian spirits of the lot become the
three Fates — Moirai in Greek — the stern divinities who sit
and spin the thread of human destiny, ordaining for every
man at birth all the events of his life, and especially the last?
This will be the subject of the next chapter.
133 Hdt. 6. 56, X. RL. 15. 3. In Kos there was a priesthood with the title
yepcapSpos pcroiAtcov: Paton xxxv. On the royal privileges of the Kodridai at
Ephesos see below p. 545.
134 Th. 1. 13. 1. 136 II. 12, 310-21, cf. 17. 250.
IX
MAN’S LOT IN LIFE
I. Occupational Clans
In the higher grades of tribal society specialised occupations
tend to be hereditary in particular clans. In ancient Greece we
hear of many such craft clans: the Asklepiadai (physicians), the
Homeridai (minstrels), the Iamidai, Branchidai, Krontidai
(prophets), the Kerykes, Theokerykes, Talthybiadai (heralds ). 1
At Sparta all heralds were Talthybiadai. As Herodotus puts it,
heraldry was the glras of the clan . 2 And there are many more
whose names are vocational: the Poimenidai (herdsmen),
Bouzygai (ox-spanners), Phreorychoi (well-diggers), Daidalidai
(sculptors), Hephaistiadai, Eupyridai, Pelekes (armourers and
smiths ). 3 These craft clans might also be described as guilds.
The medieval guild — a professional corporation to which ad-
mission was obtained by some form of co-option — was, as
Gronbech has shown, a modified survival of the craft clan . 4
But the Greek guilds stood closer to their origin. In early
times the Homeridai had been actual descendants of Homer;
it was only later that they admitted minstrels who had no
ancestral connection with the founder.® The birthright was
extended by co-option. The Asklepiadai opened their ranks in
the same way, and in their case we know how it was done. The
new member swore ‘to show the same regard for his -teacher
as for his parents, to make him his partner in his livelihood, to
share his earnings with him in time of need, and to treat his
1 Roschcr LGRM s.vv. Asklcpios, Branches, Hsch. KpovrlScn, Geok^pukej.
On the Kcrykcs see above pp. 127-8, Toepffer AG 80-92. There were
branches of the Iamidai in Elis,. Sparta, Messenia, and Kroton: Hdt. 9. 33,
5. 44. 2, Paus. 3. 12. 8, 4. 16. x, 6. 2. 5, 8. 10. 5, Pi. O; 6. On modern
craft clans sec Hollis NLF 8-1 1, Landtman 83.
2 Hdt. 7. 134.
3 Toepffer AG 136-46, 166, 310-5.
4 Gronbech I. 35.
6 Pi. N. 2 1 sch., Harp. ’OpripIScci, cf. Str. 645; see p. 550.
IX
man’s lot in life 333
kinsfolk as his own brothers’.® This was a form of adoption,
which, as a rite of rebirth, had been a normal feature of the
primitive clan (pp. 45-7). It is probable that, in spite of these
modifications, die hereditary strain was never entirely lost. In
the fourth century it was still possible for Aristode, who
belonged to the Asklepiadai, to claim descent from Asklepios. 7
This was doubdess because, in Greece as in the middle ages, the
son tended to follow his father's calling.
The Asklepiadai traced their ancestry to the patron of
medicine; the Homeridai to the greatest of minstrels; the
Iamidai to a son of Apollo, the god of prophecy; the Kerykes to
a son of Hermes, the god of heraldry; the Theokerykes to the
herald Talthybios; the Daidalidai to the mythical craftsman of
Minoan Crete (p. 285); the Bouzygai to Bouzyges, who first
harnessed oxen to the plough. In each case the clan attributed
its hereditary occupation to its founder.
Before making war on Kronos and the Titans Zeus swore to
the gods that, if victorious, he would not only respect existing
privileges but bestow others on those who had none. The
result was that, when the war was over, he was invited to
assume the sovereignty.® He became king in reward for
military service. After assuming power, he distributed his
honours. The geras of Hephaistos was fire;® the moira allotted to
Adas was to hold up the sky; 10 the moira of the nymphs was to
care for mortals in the days of their youth; 11 to Apollo was
assigned music and dancing, while lamentation was the lacbos
of Hades. 12 Once Aphrodite, whose moira or timl was making
love, was caught working at a loom, whereupon Athena pro-
tested that, since Aphrodite had stolen her kleros, she would no
longer pursue the vocation which the Moirai had assigned to
her. 13 When Apollo rescued Orestes from the Erinyes, they
accused him of stealing the lachos which they had received
from the Moirai at birth. 14 Asklepios was punished for the
® Hp. Jusj. 1. 298-300 Jones. It is not expressly stated that this was
the oath of the Asklepiadai, but I do not see what other organisation it can
be referred to.
7 D.L. 5. 1.
* Hes. Tb. 73-4, '112-3, 383-403, 881-5, cf. A. Pr. 218, 244-7, Alcm. 45.
9 A. Pr. 38. 10 Hes. Tb. 520. 11 Hes. Tb. 348. 12 Stes. 22.
13 Hes. Tb. 204-5, Nonn. D. 24. 274-81. 14 A. E. 173. 335-6, 730.
IX
334 STUDIES IN ANCIENT GREEK SOCIETY
same reason: in seeking to raise die dead he had trespassed on
the moira of Hades. 16
According to Herodotus, it was Homer and Hesiod who
‘created the Greek theogony’ and ‘gave the gods their titles,
distinguished their privileges and crafts, and fixed their form’. 16
Invading tribes had overrun the Aegean, and so the Sons of
Kronos had conquered the world. The invaders had divided the
land by lot, and so the Sons of Kronos divided the world. The
kings of these tribes owed their position to military service,
and so did the King of Olympus. In the same way the mythical
division of labour among the gods reflects the system of oc-
cupational clans, a system in which a man’s vocation — his
portion in life, his birthright — had been determined for him
by the clan into which he was born.
2. The Moirai as Spinners
How did these Moirai — ‘portions* of wealth or ‘divisions’ of
labour — become the three spinners of destiny? In looking for
the answer, we must try and do better than Wilamowitz, who
treated the idea as ‘a mere poetical invention', as though
poetical inventions were either self-explanatory or inexplicable. 17
Three was a magical number, associated, -among other things,
with the three lunar phases. 18 As divider of time and an object
of women’s worship, the moon was doubly connected with the
Moirai, who are always female. Their names are Klotho,
Atropos, and Lachesis. Klotho is simply spinning personified,
and the oldest of the three. Homer speaks of them collectively
as Klothes, but does not mention the other two. 19 Atropos
16 A. A. 1004-14.
18 Hdt. 2. 53.
17 Wilamowitz GH x. 359. Krause 152 attributes it to 'speculation.'
W. Drcxlcr (in Roscher LGRM 1. 2715) argued that the Moirai were
goddesses of clouds and mists, which primitive man imagined as 'a
sort of spinning’ as he watched them floating in wisps across the summer sky.
18 Briffault 2. 603-6.
19 Od. 7, 197. The trinity first appears in Hes. Th. 218. The generalised
use of lm&&dco (I/. 24. 525, Od. 1. 17) shows that the concept of KAw6es
was already ancient in the Homeric period. Another stereotyped expression
of the same kind is 6ewv Iv youvaai keItoci (II. 17. 514, Od. 1. 267) referring
to the unworked wool lying on the spinner's knees: Onians KG.
ix . - ' man's lot in life 335
appears in later literature as goddess of the abhorred shears who
‘slits the thin-spun life'. This image is apparently based on
cutting the web from the loom: 'I have rolled up like a weaver
my life; he will cut me off from the loom.' 20 But it is not found
in early Greek literature, 21 nor does it follow from the tradi-
tional interpretation of the name — she who cannot be turned
back, whose thread cannot be unspun. 22 And even this interpreta-
tion, which goes back to AEschylus, seems to be in the nature
of an afterthought. It is not hard for the spinner to unwind
what she has spun or for the weaver to unravel what she has
woven: Penelope is a standing instance to the contrary. Per-
haps therefore it is a false etymology. The word is based on the
idea of turning (trfpo ) — of that there is no doubt; it may be,
however, that the prefix is not privative but intensive. In that
case Atropos is just a by-form of dtraktos, with interchange of
j> and k — not ‘she who cannot be turned* but the Turner —
a personification of the spindle. As for Lachesis, the goddess
of the Idchos or allotted portion, her place beside the other two
suggests that she again must have carried some connotation
germane to the art of spinning — either the allotment of the
unworked wool among the spinners, or, what comes to the
same thing, the amount of wool required to fill the spindle. 23
How then did this trinity become spinners of fate? The
answer must be sought in their human prototypes. We must
also observe — the tradition is insistent on this point — that a
man’s destiny is spun for him at the moment of his birth. 24
This brings them into relation with Eileithyia, who was also
pictured as a spinner. 25 From this point of view the Moirai are
the midwives, the elder kinswomen in attendance. 28 What
28 Isaial 38 . 12 .
21 1 have not succeeded in tracing this idea in ancient literature at all,
though it seems to be implied in Verg. A. 10. 814.
22 A. E. 335-6 (see my note), PI. R. 6 zoe, Call. IP. 103, Nonn. D.
25. 365, 40. 1, Luc. JTr. 18, E. fr. 491, Jo. Diac. ad Hes. Sc. 236.
23 Orph. t. 70, AP. 7. 5, Erinn. 23.
24 II. 20. 127-8, Oi. 7. 197-8, A. E. 348, E. Hel. 212, IT. 203, Ba.
99, cf. Plu. M. 637E
25 Paus. 8. 21. 3, cf. Pi. O. 6. 42.
28 Cf. Earthy 69, Roscoe B (1911) 51, BB (1923) 242, BTUP24, Hutton
233.
IX
336 STUDIES IN ANCIENT GREEK SOCIETY
then were these women engaged in spinning at the birth of a
child? To this question I can see only one answer. They were
making its clothes.
The primary function of clothes, which, at least in the
colder climates, is to protect the body, is everywhere en-
crusted with magical practices founded on the notion that a
man’s clothes are somehow bound up with his life. This ex-
plains the custom of decorating the body with magical devices,
painted or tattooed, and the wearing of ornaments, such as
necklaces, bracelets, and rings . 27 In Greece the new-born child
was wrapped in swaddling-bands and adorned with amulets.
Such articles were known collectively as gnorismata , ‘tokens’,
because they were sufficiently distinctive to identify the
wearer . 28 When an unwanted child was exposed, its tokens,
were exposed with it. This was done even when, so far from
hoping it might survive, the parents were determined it
should perish. When the infant Cyrus was handed over to a
shepherd to be abandoned for wild beasts to devour, it was
attired in richly embroidered linen and ornaments of gold;
and when the compassionate shepherd substituted for it his
own still-born child, he transferred the tokens from the one to
the other . 28 The custom of exposing the tokens cannot there-
fore in general have been prompted by the hope of subse-
quent recovery, though that may have been a secondary motive
. in particular cases. It was a ritual act, inspired by the belief that
die child’s soul was pardy contained in its clothes, which bore
the marks of its origin.
The Arabs brand their cattle with a distinctive sign called
w asm, which originally, according to Robertson Smith, was a
clan emblem, like those employed by the Bantus to mark both
27 Robertson Smith RS 3 3 5, Karstcn 1-197, Hollis NLF 2 7.
28 Roman children wore round the neck a little box containing a phallus,
the boys until they assumed the toga virilis, the girls probably till marriage:
Darcmberg-Saglio s.v. Bulla. What happened to the Greek tokens is not
so dear, but in the case of Orestes they were carefully preserved, and
in another instance they were dedicated by a girl at marriage: Longus 4. 37,
cf. p. 248. At Athens the swaddling-bands were commonly made from the
dothes, kept for the purpose, in which the parents had been initiated at
Eleusis: Ar. Pi. 845 sch.
28 Hdt. 1. 111. 3, 113, 2, cf. Horn. H. 3. 121-2.
IX
man's lot in life 337
the clan cattle and the clansmen. 80 The word w asm is cognate
with ism, the Arabic for ‘name'. The same equation — ‘mark',
'name', ‘know', 'kin' — is found in the Indo-European languages
(p. 46). The Theban Spartoi had two emblems, a snake and a
spear. The story was that every member of the clan was marked
with a spear from birth. But birthmarks are not hereditary, and
it has been plausibly suggested that the spear was really a tattoo. 31
It served the same purpose as the snake necklace with which the
daughter of Erechtheus adorned her child in memory of her
ancestor, the snake-man (p. 1 16). When Orestes returned home,
he proved his identity to his sister, who had not seen him since
he was a child, by showing him a garment she had woven for
him — probably his swaddling-bands. 32 It was embroidered with
animal designs. Throughout antiquity animals were a tradi-
tional motive in the metal ornaments and embroidered linen in
which infants were attired. There are several instances in
Menander. Syriskos is examining the tokens of a foundling:
‘Here’s an iron ring plated with gold, and on the seal is carved
— is it a bull or a goat?' Again: ‘Go and fetch the casket with
the embroideries in it — you know, the one I gave you to keep.
. . . Isn’t this a he-goat or an ox or some such beast? . . . That’s
the attire they found me in as a child.' 33 The gnorlsmata were
survivals from the time when children had been marked with
the clan totem. They signified that, as a reincarnation of the
clan ancestor, the child had inherited by right of birth the
ancestral duties and privileges — the moira — of his dan. And
hence, as projections of the women who wove the embroidered
swaddling-bands, the Moirai stood for the authority of an-
cestral custom, which determined each man’s birthright.
The same conclusion is reached by another line of argument.
The daimon of the Orphics and Pythagoreans was the genius or
guardian spirit who took charge of a man at birth and decided
all the crucial issues of his life. This is the Egyptian ha, the
30 Robertson Smith KMEA 213, Hollis MLF 290, NLF 22.
31 Arist. Poet. 1454b, D. Chr. 1. 149R, Hyg. F. 72, Plu. M. 563a;
Harrison T 435, cf. Cook Z a. 122.
32 A. C. 230 sch.
83 Men. Ppit. 170-4, PK. 631-60. In Philost. Im. 1. 26 the Horai
scatter flowers on the swaddling-bands of the infant Hermes 'in order that
they may not lack markings.’
X
IX
338 STUDIES IN ANCIENT GREEK SOCIETY
Mexican nagual, the Amerindian manitoo 34 — individual totems
formed by analogy from the clan totem, with which in many
cases they are combined . 88 And even in Greek, besides this
individual iaimon , we hear of a hereditary iatmon belonging to
the clan . 88 Moreover, this word was constantly used in such a
way as to be virtually interchangeable with moira. The Greek
for ‘trying your luck’ was alternatively ‘to put your iaimon to
the proof’ or ‘to ascertain your moira*. m Empedokles says there
are two kinds of dalmones or moirai that inaugurate a man’s life . 88
Iphigeneia cries out in the same breath against the unlucky
iaimon that brought her from the womb and the Moirai who
delivered her mother of a child so miserable . 89 And, to clinch
the matter, iaimon is cognate with dais, ‘meal’, and iasmSs,
‘division’. It is the ancestral spirit who determines each man’s
moira .
The Moirai were also active at initiation, marriage, and
death. At Athens, when a man returned home after being
reported dead and duly lamented by his kin, he was read-
mitted to the community by a ceremony consisting of a mimic
birth, and he was described as a ieuterSpotmos , 49 one who had
received a second pitmos—pStmos being synonymous with
moira in the sense of that which ‘falls’ to one's lot (Latin
casus). In myth, it was the Moirai who attended the bridal
bed of Zeus and Hera . 41 In cult, the bride offered a lock of her
hair to Artemis and the Moirai . 42 Of the bridal night it was
said, ‘This night inaugurates a new pitmos, a new iaimon \ 48
And phrases like moira thanatou, ‘portion of death ’, 44 show that
34 T. E. Peet in CAH 1. 334, Budge GE 1. 163, Moret 8, 145, Meek
202-7, Bancroft 2. 277, Schoolcraft 196.
at Webster 154.
38 A. A. 1478, 1568.
37 A. Th. 493, C. 51 1.
38 Emp. i22=Plu. M. 474b.
39 E. IT. 203-7, cf. Hcl. 212-4, II. 3. 182. The Salpcov yEvKtoo s of Pi. O.
13. 105 is the ird-uios auyysv^; of N. 5. 40.
40 Plu. M. 265a, Hsch. SompfriroTuos* Among the Hindus a man who
returns from abroad has to be ‘born again’: Frazer GB-TPS 1 13.
41 At. Av. 1731-43, Pi. fr. 30.
42 Poll. 3. 38.
43 Antipho Soph. fr. 49.
44 A. Pr. 919, A. 1463, cf. II. 16. 457, 23. 9.
IX
man’s lot in life
339
man had his allotted portion in the life beyond the grave. The
key to this complex of ideas is that birth, initiation, marriage,
death — the normal divisions or moirai of human life — are
treated in primitive thought as events of the same nature.
The Moirai originated as impersonations of ancestral cus-
tom, as symbols of the economic and social functions of
primitive communism — the sharing of game, the sharing of
booty, the sharing of land, the sharing of labour between the
clans; that is to say, they grew out of the neolithic mother-
goddesses, who, emanating from the female elders of the
matriarchal clan, symbolised the collective authority of count-
less generations of ancestresses who had held undisputed sway
over the lives of men ever since they had lived in clans.
y'Eschylus remembered that in the beginning of the world the
Moirai had been supreme. 46
3. The Horai and Charitcs
Having identified the Moirai, we have little trouble in in-
terpreting the Horai and Charites, who figure in English
poetry as the Hours and Graces.
The names of the Horai — Eunomia, Eirene, Dike 48 — belong
to class society. Eunomia, Law and Order, speaks for herself;
Eirene, Peace, was an idea that took shape with the city-states; 47
and, as we shall see immediately, Dike is a post-Homeric
substitute for Moira. But they are older than their names.
Their primitive character is revealed in their collective name,
referring to the divisions of the year, and in their worship as
spirits of fecundity. 48 Besides ordaining the times and seasons
of labour on the land, 48 besides filling their baskets with
flowers, sheaves, and fruits, 60 they assisted at the wedding of
Semele, 61 and swaddled the new-born Hermes, dandling him on
their knees. 52 The Charites are Euphrosyne (Mirth), Thaleia
46 A. Pr. 531-4. 46 Hes. Th. 901-2. 47 Hasebroek 118.
48 Philoch. 18, 171, cf. Ar. Pa. 308. 48 Hes. Th. 903, Paus. z. 40. 4.
68 Eus. PE. 3. 11. 38, cf. AP. 6. 98. .
64 Nonn. D. 8. 4-5, cf. Mosch. Id. 2. 164.
62 Pi. P. 9. 59-62, Philost. Im. 1. 26, cf. E, Be. 418-20, Paus. 2. 13.
3. They were also goddesses of childbirth: Nonn. D. 3. 381-2, 9. 12-6,
16. 396—8, 48. 8ox.
340 STUDIES IN ANCIENT GREEK SOCIETY IX.
(Revelry), and Aglaia (Radiance ). 68 They danced at the
wedding of Kadmos , 84 danced at the wedding of Peleus , 66
danced with Aphrodite , 68 danced with the Horai at Olympian
feasts , 67 danced with the Horai and Moirai at the birth of
Persephone . 68
It is evident that these three trinities are really one. They
HG. 51. Clarites: Attic relief
are all simply the anonymous plurality of ancestresses as dis-
tinct from the individualised mother-goddesses who emerged
out of them with tire rise of the archaic matriarchal state.
Again and again they appear as a sort of chorus with one or
other of the mother-goddesses as coryphaeus. Artemis was
worshipped in conjunction with the Moirai . 68 Demeter bore
the tide of Bringer of the Horai . 80 The Argive Hera had- a
83 Hcs. Th. 907-9. 64 Thgn. 1 5-6. 60 Q.S. 4. 140.
80 01 18. 193-5, cf. 8. 362-6, II. 5. 338. 67 Horn. H. 3. 194-6.
68 Orph. if. 43. «b CIG. 1444.
00 Horn. H. 2. 54, 192, 492, cf. Call. Cer. 122-4, Norm. £>. jj. 501-4,
IG. 12. 5. 893 'AtoWiAwv '(ApoptBcav, cf. Call. Ap. 87, Paus. 3. 18. 10, 8. 31.- 3.
IX
man's LOT IN LIFE' 34I
crown adorned with figures of the Horai and Charites. 81 In the
Odyssey the dance-leader of the Charites is Aphrodite. 82 In
Attic tradition the same goddess was described as the eldest of
the Moirai. 88 Really she was one of the youngest.
4. The Erinyes
The Erinyes are at first sight entirely different:
Maidens abominable, children grey with years,
With whom no god consorts nor man nor beast,
Abhorred alike in heaven and on earth,
For evil bom, even as the darkness where
They dwell is evil, the abyss of Tartarus. 04
Their special concern was to punish the homicide of a kinsman,
perjury, unfilial conduct, and inhospitality. The first of these
offences has been discussed in Chapter IV. The others cor-
respond to the three ‘unwritten laws* of the Eleusinian and
Orphic Mysteries: Honour the gods, honour thy parents,
honour the stranger. 06 The penalties they applied were insanity,
famine, sterility, pestilence. Originally they were believed to
act immediately, causing the offender's death. That explains
their part in the ordeal by oath, which was said to have been
instituted by Rhadamanthys, the legendary lawgiver of
Minoan Crete. 00 The accused uttered a conditional curse on
himself— he prayed that, if guilty, he might perish together
with his clan (p. 134). Accordingly, when the oath ceased to
be an actual test of guilt and came to be simply a means of rein-
forcing the evidence by intimidating the witness, 07 the Erinyes
withdrew into the other world, where they tormented the
souls of the damned as infernal ministers of Persephone.
Their connection with the unwritten laws is already apparent
in the Homeric poems, but these laws are not primitive. The
ordeal by oath is found only in the higher stages of tribalism.
Filial obedience presupposes the family. The sanctity attaching
to suppliants and strangers was designed to meet the shortage
of man-power on the land (p. 133) and later the interests
of trade. This, together with the references to Rhadamanthys
04 Paus. 2. 17. 4. 02 Oi. 18. 193-5. 08 Paus. 1. 19. 2. 04 A. E. 68-73.
06 G. Thomson AO 1. 51-2, 2. 269-72. 08 PI. Leg. 948. 07 Diamond 52.
IX
34 ^ STUDIES IN ANCIENT GREEK SOCIETY
and Persephone, suggests that the unwritten laws were a legacy
from Minoan Crete. 88
The Erinyes figure prominently in the myth of CEdipus, who
belonged to the House of Kadmos. From his father Laios he
and his sons after him inherited a curse, embodied in the
Erinys, which eventually destroyed the dynasty. 69 That was the
Delphic tradition. 70 In the Odyssey, on the other hand, the Fury
who afflicts CEdipus is not his father’s but his mother’s, 71 and
there are other indications, besides their sex, that these spirits
belonged originally to the female line. 72 They were invoked by
Althaia against her son, Meleagros, for the murder of her
brother, and the crime for which they persecuted Orestes and
Alkmaion was matricide. 78 These were all cases of murder
within the kin-r-the most terrible sin that a tribesman can
commit. Consequently, when we find the Erinyes described as
‘curses’ 74 and imagined as snakes, we can see in them at
bottom a particular aspect of those same matriarchal ances-
tresses whose malediction upheld the inviolability of tribal
custom.
The Erinyes too were associated with goddesses, but only
with Demeter and Persephone. In Homer they share with
Persephone the duty of punishing the souls of perjurers. 76 The
Arcadian Demeter Erinys bore their name, 76 and Demeter was
everywhere associated with the snake (pp. 118, 253). The
Erinyes are the Moirai of Minoan Crete.
5. The Indo-European Origin of the Moirai
Several attempts have been made to find an Indo-European
etymology for the word erinfs, but without success. In dealing
68 The ordeal by oath is more prominent in the Laws of Gortyna than
in any other primitive code: Diamond 364-5. The Dorian conquerors of
Crete took over many institutions from the older population, who con-
tinued to observe the laws of Minos: Arist. Pol. 1271b.
60 Pi. O. 2. 42-6, A. Th. 710-2, 751-2, 770-2, 776, S. OC. 1434.
70 Hdt. 4. 149.
71 01 . 11. 279-80.
72 Like the Moirai, they were worshipped exclusively by women: E. Mel.
Capt. 18-21.
78 ll. 9. 565-72, Apld. 1. 8. 3, 3. 7. 5. 74 Sec p. 136.
76 11 9 - 454 - 7 - 76 Paus. 8. 25. 4. '
ix man's lot in life 343
with Greek, which includes a large element derived from alien
languages for the most part? unknown, something more than a
bare linguistic possibility is needed to establish an Indo-
European etymology, because from the nature of the case we
are unab le to estimate alternative possibilities on the other
side. All that can be said in the present instance is that the
word connotes ‘madness', as shown by the Latin Furia and the
Greek erinyo ‘rage *. 77 This idea is radical, and no Indo-European
origin has been found to cover it.
Moira , on the- other hand, is definitely Indo-European, and
this point enables us to press our analysis a step further. Being
associated with mother-right, die Moirai must go very far
back into Indo-European prehistory. The same conclusion
follows from their part in the sharing of meat, which takes us
back ultimately to a hunting economy. That being so, we
expect to find cognate concepts in other branches of Indo-
European culture. The subject is too wide to be investigated
here, but a few words may be said about the Roman Parc®, the
Celtic Matres Dese, and the Germanic Norns.
The Parcs do not help much. As spirits of childbirth, they
are no doubt of the same ultimate origin, but their presenta-
tion as a trinity.of spinners is due to Greek influence, and apart
from this there is Kttie to distinguish them from the host of
good and evil spirits, personifications and abstractions, out of
which the Roman pontiffs elaborated an all-embracing system
of spells and incantations designed to keep the populace in
permanent fear of the wrath to come . 78 If the idea of Moira
survived at all in Latin, it is rather to be looked for in such
words as coro 'flesh' (Umbrian karu 'portion'), sors ‘lot’. ( sero
plait'), and casus (cf. Gk. pStms).™
The Matres Dese appear on hundreds of votive reliefs and
plaques, dating mostly from the second century A.D., which
have been found in northern Italy, France, Spain, Britain, and
Germany west of the Rhine . 80 One of the types is a trinity of
seated women with baskets of fruit in their laps; another is a
»
77 Paus. 8. 25. 6.
78 Roscher LGRM j.v. Indigitamenta, cf. Plb. 6. 56. 6-12.
79 Buck CGGL 49, Emout-MeiUet j.vv.
8 ® M. Ihm in Roscher LGRM 2. 2464-79: fig. 52.
IX
344 STUDIES IN ANCIENT GREEK SOCIETY
chorus of women dancing with joined hands. In some of the
seated groups one of the figures holds a cornucopia — the Greek
horn of Amaltheia . 81 The last feature is of special interest,
because it recalls the Venus of Laussel — a paleolithic rock-
carving of a naked woman holding a bison's horn . 82 The
HG. 52. Matres Deal: relief
from Avigliano
FIG. 53. Venus of
Laussel: paheolitlic
carving
Matres Dee may be regarded as a fusion of die Celtic mother-
goddesses with older cults associated with the female figurine.
The Norns have left no cult monuments, but they figure
prominently in myth, and their resemblance to the Moirai is
very close. They too are goddesses of birth, marriage, and
death; they too are spinners of fate . 88 Greek influence can be
ruled out. It could only have come through Rome, and the
Roman concept of the Parcae as spinners, being merely a literary
borrowing, was confined to the educated classes. If it had made
any impression on popular thought, it would have left its mark
on the Matres Deal, but, though the finds are plentiful, there
appears to be only one in which these are identified with the
Parcae and none in which they are spinning. It seems clear that
the Norns and the Moirai belong to a common Indo-European
heritage.
81 See p. 250 n. 10.
82 GAH Plates 1. 8: fig. 53. -
83 Paul 3. 282, Mannhardt GM 576, 609, cf. Chadwick GL z. 208,
2x8, 646. Whether the concept of Moira occurs in other IE languages
is more than I can say, but it is worth noting that Gk polpa (1) 'share'
(2) ‘esteem’ (3) ‘fate* corresponds to the Irish (1) cion (a) 'share' (b) 'affection'
(2) einneambainl 'fate'.
IX
man’s lot in life
345
6. The Transformation of Moira
The Moirai and Erinyes arc closely related in Greek mytho-
logy. j'Eschyius says that the world was governed in the be-
ginning by ‘the threefold Moirai and the unforgetting Erinyes’,
'who were then more powerful than Zeus. 84 The women of
CEdipus cry out against ‘Moira, giver of evil, and the ghost of
CEdipus, the black Erinys'. 88 When Agamemnon repents of
having robbed Achilles of his gtras, he attributes his blunder
to the malice of Zeus, Moira, and Erinys. 80 When Zeus warns
Poseidon not to trespass beyond his moira, he reminds him that
the eldest brodicr’s portion is protected by die Erinyes. 87
In post-Homeric poetry Moira is often replaced by Dike.
When Agamemnon and Mcnclaos refused die right of burial —
die moira of the dead — to Ajax, the dead man’s brodier cursed
diem in the names of Zeus, Erinys, and Dike Svho brings
fulfilment’ ( telesphSros ). 88 This was a traditional epithet of
Moira. In die Oresteia parents struck down by their children
appeal to Dike and the Erinyes, and in the same way Hcrak-
ieitos declared diat, if the Sun were to transgress his appointed
'measures’ ( mltrd ), lie would be found out by the Erinyes,
ministers of Dike. 88 The idea of mitron was a post-Homeric
development of moira. to These passages show that Dike has
taken over the functions of Moira, 81 and that both are related
to the Erinyes. The nature of die relationship seems to be
that, whereas Moira is offended by transgression of die limits
sec to human conduct, die actual chastisement of the offender
is left to the Erinyes. The Moirai decree, die Erinyes execute. 88
And diis traditional co-opcradon corresponds to the fusion of
cultures underlying Greek civilisadon, the dominance of the
Indo-European element being reflected in the superior
authority of the Moirai.
84 A.Pr. 531-4. 88 A. Tb. 962-4. 88 1 /. 19. 86-7. 87 Ib 15.204.
88 S. Aj. 1389-92, cf. 1326-7, An. 1070-5, A. Pr. 527.
88 A. E. 514-5, Heraclit. 94. 88 G. Thomson AA 78.
81 Cf. Pi. P. 4. 145-6 with PI. Leg. 943c, Hes. Op. 256-62. So the
Homeric Kerri yoTpav ‘appropriately’ anticipates die later Korri SIkijv; ll. 1 .
286, Hdt. 7. 3 5. 2 etc.
82 That is why it is the Erinyes who silence the horse of Achilles after it
has said all that it is fated to say: ll. 19. 418.
IX
346 STUDIES IN ANCIENT GREEK SOCIETY
In excusing his blunder Agamemnon coupled with Moira and
Erinys the name of Zeus. How did Zeus stand to the Moirai?
We turn to the evolution of the kingship. As a specialised
occupation, it- became eventually hereditary , 63 but for a long
time the succession remained subject to popular ratification.
When the sons of Temenos murdered their father in order to*
secure the succession, the people intervened and gave it to the
son-in-law . 04 Telemachos aspired to his father's kingdom, but
all he claimed as his by right was the inheritance of his pro-
perty . 66 Similarly, when Agamemnon was accused of taking
more than his share of the spoils , 66 we can see that the king
was beginning to claim as a right what was properly a gift
from his people. So with Zeus. In the beginning, according to
AEschylus, he had been powerless to override the Moirai . 07
When his son Sarpedon was about to fall in battle, he was
tempted to save him, but Hera deterred him with the warning
that, if he violated the decrees of fate, other gods would follow
his example . 08 On the other hand, such expressions as moira
than and epekUsanto theol are signs that the Moirai are sur-
rendering their authority , 66 and their eventual subordination
is revealed at a later date in the cult-title moiragetes, applied
to the Olympian Zeus and the Delphic Apollo . 100 The new
gods have conquered. The tribe has been superseded by the
state.
Spinning was the woman's task. The significance of Moira
from this point of view may be contrasted with another
element in Greek thought, which had its origin in the work of
the men. The notion of pasture underlies a word which in
social importance was eventually to eclipse Moira. The word
nStnos denoted originally, like moira itself, a 'division' or
‘portion’, but it differed in two respects. It had no connection
with the lot, and it was applied only to pasture . 101 Long after
the moira of the clan had been broken up into family holdings,
03 Cf. Roscoe B (19x1) 13, Radin 16. Apld. 2. 8. 5; see p. 166.
06 Od. 1. 389-98. 00 II. 9. 330-4, 367-8. 07 A. Pr. 534.
08 II. 16. 433-49. 00 Od. 11. 292, 22. 413, 1. 17, II. 24. 525.
100 Paus. 5. 1 5. 5, 8. 37. i,io. 24. 4, cf. 1. 40. 4, E. fr. 260, El. 1247-8,
Mel. fr. adesp. 5, Orph. H. 59. 11-4, fr. 248. 4.
101 Cornford RP 27-31.
IX
man’s lot in life 347
the pastures remained common, their use being regulated by
customary rights. In this way tt6mos developed the sense of a
common usage, an accepted custom, and so custom as by law
established. 102 Both had their roots in tribal life, but, whereas
at the opening of the historical era Moira was already on the
wane, Nomos only matured in the democratic city-state.
The decline of Moira and the rise of Nomos mark the transition
from the matriarchal tribe to the patriarchal state.
With the growth of class inequalities, the use of the lot for
the distribution of wealth became more and more restricted,
with the result that the Moirai, who had asserted the birthright
of all men to the fruits of their labour, were transformed into
inexorable Fates whose authority was used to reconcile men
with their lot, however meagre, in the new social order, in
which the majority had been dispossessed; and consequently,
robbed of their birthright in the real world — their share in 'the
fatness of the earth and plenty of corn and wine’ — they were
driven to console themselves with the mystical hope of re-
covering their lost heritage in an illusory world beyond the
grave. The birthright became a deathright.
ioa cf. flflos (i) 'habitat' (2) ‘habit’, 'custom'.
X
THE FORMATION OF TOWNS
i. Thucydides on Primitive Greece
In his exposition of the origin and growth of die Greek pSlis
Thucydides reveals himself as a materialist historian of the
first rank:
It is apparent that the country now called Hellas was in early times in-
securely occupied and subject to frequent movements of population as each
community found itself dispossessed by ever more powerful invaders. In
the absence of trade and free intercourse by land or sea every community
depended on its own territory for the necessities of life. There was no super-
fluity of wealth, and no cultivation of fruits, because they never knew when
they might be dislodged from their unfortified habitations; nor were they
reluctant to move, being confident of obtaining anywhere the means to
satisfy their immediate needs; and hence their towns were not distinguished
by size or other marks of power. Such migrations were always most frequent
where the best soil was to be had — in Thessaly, Boeotia, the greater part of
the Peloponnese, excepting Arcadia, and other fertile districts. The fruitful-
ness of these regions favoured the accumulation of wealth, which promoted
in turn destructive civil factions, thereby inviting foreign attack. Attica, on
the other hand, preserved from discord by the poverty of her soil, remained
in uninterrupted possession of the same people. . . .
It is dear that before the Trojan War the country had never joined in any
common enterprise or even possessed a common appellation. The name
Hellas dates only from Hellen. Before his time there had been separate
names for the several nations, of which the most considerable was the
Pclasgian. ... At that time the various peoples of common speech that came
eventually to be called Hellenes were too weak and scattered to act in con-
cert, and it was only after making progress in maritime communication
that they engaged in the Trojan War.
The earliest ruler known to have possessed a fleet was Minos. He made
himself master of the Greek waters and subjugated the Cyclades by expelling
the Carians and establishing his sons in control of the new settlements
founded in their place; and naturally, for the safe conveyance of his revenues,
he did all he could to suppress piracy. The early Greeks, like the barbarians
of the islands and coasts, had taken to piracy as soon as they learnt to sail
the seas. Commanded by capable leaders, for the sake of personal gain and
the relief of their poor, they raided and ravaged die unfortified groups of
village communities, gaining the greater part of their livelihood in this way.
X
THE FORMATION OF TOWNS
349
In those days, so far from being discreditable, the occupation was rather a
source of honour, as it still is, subject to the observance of certain decencies,
among some peoples of the mainland; and that is how it was regarded by the
early poets, who represent mariners as being asked on landing whether they
are pirates, implying that the truth would not be denied and that there was
no reproach in the question. Forays were also undertaken by land, as is still
done by the Lokroi Ozolai, Aitoloi, Akamanes, and other Greek peoples. . . .
In regard to towns, those founded in the later period and possessed of some
surplus wealth were all fortified and situated by the sea on narrow necks of
land in order to facilitate trade and strengthen their hold on their neigh-
bours; but the earlier foundations, being exposed to the attacks of pirates,
were located at a distance from the sea. 1
Such unstable conditions placed a premium on pastoral
wealth, which had the advantage that, when' the people were
forced to move, they could take their flocks and herds with
them. The Homeric poems contain many allusions to cattle-
raids, 2 * * * * * and the marauding expeditions described by Thucydides
were no doubt largely of this character. The same conditions
had a contrary effect on tillage. Vines and olives in particular
were a long-term investment requiring irrigation (p. 310)
and in early times they were precluded for this reason. The
instability of the settlements did not permit of a permanent
attachment to the soil. And so in the more backward areas
agriculture was restricted to cereals, which yielded a quick
return. The state of the country was therefore favourable to
migratory agriculture, providing for periodical redistribution
of the arable. 8
Thucydides applies the term pSlis to both types of settle-
ment — the unfortified group of villages and the fortified town.
This comprehensive use of the word argues an unbroken
development from the primitive village community to the
imperial city whose downfall he lived to see.
2 . Formation of Towns in the Historical Period
The difference in cultural development between town and
country is a familiar feature of modern capitalism, in which the
1 Th. x. 2-5, 7. 2 II. x. 154, ii. 672, 20. 91, Od. 21. 16-9.
8 Cf. Das 445, writing of the Kuki tribes of the Manipur Hills: ‘Shifting
cultivation does not help the accumulation of wealth in individual hands
and the consequent growth of rank; on the other hand, it has bred an
extreme democratic spirit in their social and political life.’
X
35° STUDIES IN ANCIENT GREEK SOCIETY
interests of the countryside are subordinated systematically to
those of the town. The industrial town is the residential unit
characteristic of capitalist production; the village is a remnant
of feudalism.
In ancient Greece a similar difference was apparent, but, in
keeping with the lower level of society as a whole, it manifested
itself as a contrast between civilisation and barbarism. The
city-state was a recognised mark of civilised life as opposed to
the barbarous or semi-barbarous custom of living in open
villages. Strabo repeatedly draws attention to the village life
of the barbarians:
Some writers state that the Iberians of Spain have over a thousand towns,
but I think they must be referring to their large villages. Whether it is
poverty of soil, or the great distances, or the wildness of the country, the
natural conditions do not permit of many towns, and, apart from the
southern and eastern coasts, the life and manners of the people suggest
nothing of the kind. The majority of the Iberians are village-dwellers, and as
such they are uncivilised. 4
He is well aware that the Greeks themselves had once lived in
the same way:
In Homer's time the present city of Elis had not been founded; the people
lived in villages. ... It was not till after the Persian Wars that the city came
into being, formed by the federation of several demes. The same is true of
nearly all the Peloponnesian settlements, with the few exceptions that the
poet names. They were not cities but regional groups of demes, from which
the cities we know afterwards arose. Mantineia, for example, was a com-
bination of five demes, Tegea and Heraia of eight, Patrai of seven, Dyme of
eight. 6
Similarly, the territory of Megara had consisted of five dis-
tricts, each comprising several villages, which were eventually
united under the city of that name.® In many cases the
urbanisation was effected by stages — the combination of
villages into a town, the combination of towns into a city. In
Roman times, for example, the island of Keos contained two
towns, Ioulis and Karthaia, but previously there had been
four. 7 The unification of Rhodes^ated from 508 B.c. The new
capital, named after the island, drew its population from the
4 Str. 163, cf. 186, 218, 241, 250, D.S. 5. 6, Hdt. 1. 96.
6 Str. 336-7, cf. PIb. 4. 73. 7. 0 Plu. M. 295b, Th. 4. 70.
7 Str. 486, cf. 657, Paton xxvii.
X THE FORMATION OF TOWNS 351
three ancient towns of Lindos, Ialysos, and Kameiros; and
since each of these contained a number of d ernes, it is clear
that they too had evolved from groups of adjacent villages. 3 * * * * 8 *
These examples show that the making of city-states went on
all through Greek history. In Thucydides' day the Greeks of
Aitoh'a and Akamania were still living in open settlements, 8 and
even Sparta was only a group of villages 'after the old Hellenic
fashion'. 10 A generation or so later two events provided as it
were a practical demonstration of the city-state’s anatomy. In
386 B.c. the Spartans forced the citizens of Mantineia to
demolish their city and scatter themselves among the villages
out of which it had been formed. 11 They were trying to put die
clock back. Twenty years later, when Spartan rule had been
overthrown, the disjecta membra of Mantineia were reunited,
and in addition a new city was founded at Megalopolis, in-
corporating all the towns and villages for miles around. 12 We
see what Aristotle meant when he defined the pdlis as 'a union
of several villages'. 13
3. From Tribal Camp to City-state
In Chapter VIII it was argued that the Attic demos originated
as a village community of a familiar type — the territorial unit
corresponding to the clan. The current Greek for ‘village’ was
hSme. Aristotle says expressly that kSme (kSma) was the Doric
equivalent of the Attic demos; and just as Plutarch couples deme
and clan together, implying that they had once been co-
extensive (p. 326), so in another passage Aristotle describes the
city-state ( pdlis ) as ‘a union of clans (j>/ne) and villages ( \omai )’. 14
The typical Greek village was in origin a clan settlement.
The history of the pdlis is imbedded in its name. The upper-
most stratum in the meaning of the word is 'city-state' in the
8 Str. 653-4, SIC. 339 n. 2, 570 n. 4. 9 Th.3. 94. 4.
18 Th. 1. 10. 2. 11 X. Hell. 5. 2. 5-7. 12 Paus. 8. 27, D.S. 15. 72.
13 Arist. Pol. 1252b. In SIC. 344 (Hellenistic period, Teos) we have
a decree for rebuilding the town with the object of absorbing the population
of Lebedos, who are to be known henceforward as Teioi, thus losing their
separate identity, except that they will have a cemetery of their own.
14 Arist. Poet. 1448a, Pol. 1281a. 14.
X
352 STUDIES IN ANCIENT GREEK SOCIETY
abstract — the organisational form characteristic of civilised
society. Beneath this lies the concrete application to the city
itself, including its satellite villages. At this level a distinction
emerges between the citadel and the lower town, the akrSpolis
and die dsty. The former term, universal in later Greek, occurs
twice in the Odyssey, but in the Iliad we find only the dis-
junctive form dkre p6lis, and in several passages the citadel is
simply the p6lis without qualification. 16 This brings us down
to the basic meaning of the word — a natural stronghold, such
as would have been a good site for a village setdement in the
troubled times described by Thucydides.
Let us now try to sketch the growth of these Greek villages,
guided by the general considerations advanced in the preceding
chapters and availing ourselves where possible of additional
evidence bearing on points of detail. Our picture will neces-
sarily be over-simplified, but it will give us a working hy-
pothesis which we can then proceed to test.
The earliest Greek tribal setdements were temporary. They
may have been occupied for a year, for several years, for a
generation, but sooner or later the tribe was forced to move. It
is clear that a shifting setdement of this kind must have re-
produced the plan of the nomadic tribal camp, because, except
to the extent that it has become fixed, it is nothing but a camp.
The evidence relating to tribal camps is plentiful, though it
has never been collected. I will cite some typical examples.
The Australian Arunta arrange their camp in a circle, which is
divided into two semicircles, one for each moiety, and four
quarters, one for each phratry. Members of a phratry are at
liberty to pitch their windbreaks anywhere within the allotted
quarter, except that a patch of ground is reserved as a meeting-
place. 16 The Amerindian camps are planned on the same
principle. The Kansas tribe, for example, is divided into two
phrattics of eight clans each, and their camp takes the form
of a circle bisected by the line of march, each phratry being
accommodated in one of the semicircles. One way of stating
the rule of exogamy is to say that a man must take a wife from
16 01 8. 494, 504, 11 . 6. 88, 257, 297, 317, 22. 383 etc.; ll 6. 86,
17. 144, Oi. 14. 472-3.
16 Spencer A 501-4.
X
THE FORMATION OF TOWNS
353
the far side of the circle. 1 ’ The tribal camp is thus a diagram of
the tribal system. So, when the tribe has ceased to be nomadic,
the village settlements reproduce the same pattern. Each clan
resides in its own village or ward. 1 8 In Mexico the land was
divided among the clans ( calpulli ) subject to annual reallotment,
and the townships in which the people resided were divided
into as many wards as there were clans. 19
We have seen that in Rhodes the three ancient townships cor-
responded to the three tribes (pp. 324-5) and were surrounded
by villages corresponding to the clans (p. 326). Early Rhodes
was thus a typical tribal settlement. Similarly, the Aitoloi, who
in the fifth century were still living in open villages, were
divided into three tribes — the Apodotoi, the Eurytanes, and
the Ophioneis. 29 The people of Malis were also divided into
three, and the island of Zakynthos was a tetrapolis or confederacy
of four towns. 21 A similar origin may be postulated for the
ancient Attic tetrapolis of Marathon, Oinoe, Probalinthos, and
Trikorythos, which was maintained as a religious union. 22
These are all tribal confederacies at different stages of growth or
decay.
Let us try to envisage the manner in which such settlements
were formed. This means going back to the neolithic village
community. To simplify the argument I shall assume that the
settlers are already patriarchal, although in neolithic Greece
this would have been the exception rather than the rule.
The territory to be occupied by the tribe and the sub-
divisions allotted to the clans are chosen with regard to the
nature of the soil, the accessibility of water, and perhaps also
the needs of defence. Each clan centres its village so far as pos-
sible on some rocky height or natural place of vantage. The
best site is assigned to the clan of the tribal chief. If it is large
enough, the whole village is built on it; if not, the chief and his
family establish themselves on the chosen centre, with the
other households clustered round and ready in time of danger
to take refuge in his compound. Or it may happen that the
17 Dorsey 230-2, cf. 216-21, 23 3, Hale 184, Haeckel 457-60.
18 Gatschet 1. 1 54, Bourke 229. 19 Bancroft 2. 226-7.
29 Th. 3. 94. 4-5. 21 Th. 3. 92, 2. 30.
22 SIG. 541 n. 1. Its founders may hare been the Lapithai (p. 264).
Y
X
354 STUDIES IN ANCIENT GREEK SOCIETY
whole village is surrounded with a palisade or rampart, and it
may then serve as a refuge for the' other villages . 28
Within the village each household has its own residence,
with an enclosed garden attached, and somewhere within the
area, perhaps in front of the chief’s house, there is an open
space — the ancient agora, the modern platela, our own village
green. The village tree in the middle of this space, surrounded
it may be by a stone bench for the elders, is still a familiar
sight in- Greece . 24 This is where the clansmen assemble. The
heads of households enjoy the right to meet, eat and deliberate
in the chief's house. As head of the principal household, the
chief is in a special sense the representative of the clan ancestor,
and so his hearth is invested with a special sanctity. It is the
ancestral hearth of the clan, recalling the time when, not yet
divided into families, the clan had gathered round a single
camp fire. Outside the village, but within its territory, lie the
ploughed fields and beyond them the grazing land and waste.
Tillage and pasturage are controlled by the village assembly
under the direction of the elders. The arable is distributed
among the families in small open holdings and redistributed
at regular intervals. The pasture is undivided. The only well-
established division of labour is between the sexes.
So much for the internal organisation of the village. The
relations between the villages are regulated on the same
principles. Just as the house of the clan chief is the best-
protected in the village, so the village of the tribal chief is the
best-protected in the territory. Just as the clan chief presides at
the ancestral heardi of the dan, so the tribal chief presides at
the ancestral hearth of the tribe. Just as the clan chief enter-
tains the other heads of households, the elders of the dan, so
the tribal chief entertains the other dan chiefs, the elders of
the tribe. Just as each village has a meeting-place for the dans-
men, so the tribesmen of all dans meet in the principal village
to determine, under the direction of the tribal chief and elders,
major issues affecting the whole community, such as peace and
war, migration, and the reception of strangers.
23 Tritsch SA 70, cf. Baden-Powell 67.
24 Hdt.4. 15.4, cf. Baden-Powell 23, Gurdon33, ChadwickGL 1. 324. On
the relation of the agora to die king’s house sec Tritsch AE, Wycherley 21-2.
X THE FORMATION OF TOWNS 355
So for we have treated the community as a self-contained
unit. It may have relations with other similar communities,
but these are not such as to disturb its structure as a whole,
based on neolithic self-sufficiency. A new stage is reached with
the introduction of metals. Metal tools are far more efficient
than tools of wood or stone. They may be produced locally, or
they may be brought in by itinerant smiths. In either case the
effect is to. raise the productivity of labour to a level at which
it becomes possible to support a number of specialists — the
smith, the mason, the tanner, and so on; and of course these
new divisions of labour open the way in their turn for further
improvements in the technique of production. In each com-
munity there now dwell, in addition to the chiefs and culti-
vators, a number of artisans — Aemiourgoi in Greek, men who
‘work for the community', implying that they are remunerated
by the community with payments in kind. 2 6
The development of these new techniques requires, not
merely that there should be some surplus available from agri-
culture, but that this surplus, which has been scattered
hitherto in small fragments among the individual cultivators,
should be concentrated so as to make it effective. This is done
by placing it in the hands of the chiefs. The chiefs become the
recipients of regular tribute in the form of tithes or labour
services. Such payments are given freely by the clansmen as a
due return for benefits received, whether these benefits be real,
such as protection from marauders or successful leadership in
war, or imaginary, such as a plentiful harvest or some other
good fortune attributed to the magical powers of the chief. But
the underlying factor is economic. Only in this way can suffi-
cient capital be accumulated for the new technical develop-
ments. Even so metal remains scarce. The chiefs tend to
reserve the available supply for their own use. Working with
copper or bronze, the artisans invent for them a new implement,
the sword, with which they appropriate forcibly the surplus of
neighbouring communities. 86 Warfare becomes an industry.
How the lion’s share of land and loot goes to the chiefs has
been described in Chapter VUI. Each of them has now his
t imenos, tilled for him by slaves taken in war, and his surplus
26 OA. 3. 432-5, 17. 383-5, 19. 135. 26 Cf. Childe SBS 48.
X
356 STUDIES IN ANCIENT GREEK SOCIETY
is now so ample that he can engage in the production of
articles for exchange.
The exchange of commodities grew imperceptibly out of
intertribal hospitality. The laws of social intercourse evolved
within the tribal circle of kith and kin had at first been
restricted to that circle. Outside all men were strangers,
enemies, to be killed or robbed at will . 27 And hence, when
peaceful relations came to be established between tribes,
die stranger was admitted into the circle of kindred by the
symbolical act of sharing food with him. ‘Those who sit and eat
together’, Robertson Smith remarks, ‘are united for all social
effects; those who do not eat together are aliens to one another
without fellowship in religion and without reciprocal social
duties .’ 28 The stranger who has been entertained to a meal is a
stranger no longer. The enemy has become a guest (Latin
fastis). When Odysseus arrived as a suppliant at the court of
Phseacia, the first thing the king did was to give him a seat at
table . 29 In the same way, when the guest departs, he takes with
him a gift from his host as a pledge of their new relationship.
And so the exchange of gifts comes to be recognised as a
guarantee of friendship. Each parting gift constitutes a claim
on the recipient to be honoured at some future date. When
Athena, disguised as a sea-captain, says goodbye to Telemachos,
he presents her with an heirloom, which she promises to repay
on her return with a gift of greater value . 30 During the Trojan
War, Euneos called at the Greek camp from Lemnos with a
cargo of wine. After presenting a thousand measures to
Agamemnon and Menelaos he offered the remainder to all and
sundry in exchange for metals, oxen, hides, and slaves . 31 Here
we have a straightforward act of barter, but preceded by a
formal presentation. Acceptance of this is a pledge of security
for the transaction that follows. It is at the same time a per-
quisite for the kings — an impost on trade.
27 Hence yveonfc (x) ‘known* (2) 'kin*, cf. Eng. ‘kith and kin*.
28 Robertson Smith RS 269. 20 Od. 7. 167-71.
30 Od. 1. 309-18, cf. II. 6. 230-1; Bancroft 1. 192: ‘Even their system
of presents is a species of trade, the full value of each gift being confidently
expected in a return present at the next festive occasion.*
31 II. 7. 467-75, cf. the Sumerian ttigha: Langdon in CAH 1. 378.
X
THE FORMATION OF TOWNS
357
This explains why barter develops under the chief’s control.
When the community has reached the stage of absolute surplus
production, the meeting-place in die central village becomes a
market-place, where the local chiefs barter their surplus for
the surplus of other communities. The same course is followed
by the artisans. So long as their productive capacity is ex-
hausted by the needs of their fellow clansmen, their opera-
tions are confined to their own village; but when improved
technique enables them to produce more than the village
requires, they too take the surplus to market. And the result
is that all social relationships are transformed. The owner of
the tfmenos, the recipient of tithes, can no longer pretend that
his privileges are a return for services rendered, because he is
now using them to make a further profit, and this gives him
the power and the incentive to intensify the rate of exploita-
tion. So with the artisans. Once embarked on commodity pro-
duction, they find it pays to deal with their fellow villagers too
on a commercial basis. They cease to be ‘workers for the com-
munity* and become workers for themselves. The upshot is
that sooner or later both chiefs and artisans quit their villages
and set up house near the market. The central village becomes
a market town. The village handicrafts do not entirely dis-
appear, but the village is no longer self-sufficient. It becomes
increasingly dependent on the skilled labour of the town. And
in taking this step they sever their clan ties. The chiefs have
ceased to represent the separate interests of their clans. They are
becoming a landed aristocracy united against the poorer clansmen
by a common interest of class against class. The artisans on their
part organise themselves in guilds formed on the pattern of the
clan; but, so long as the economy remains agrarian, they are not
in a position to dispute the supremacy of the landowners.
As for the smallholders, the country cousins, they pursue
with their more limited resources the same objective. The very
existence of the timinos is proof of its economic superiority. Not
being liable to redistribution, it can be permanently enclosed,
with better protection for the crops, and labour can be put
into it as a long-term investment. As the open fields become
exhausted, tillage pushes out into the waste, and instead of
falling under communal control the land thus reclaimed
X
35$ STUDIES IN ANCIENT GREEK SOCIETY
becomes individual property. Meanwhile increase of popula-
tion is reducing the size of the lots in the open fields at each
redivision, and so the practice of redivision is abandoned, And,
when the holdings have become fixed, their discontinuous
distribution in strips, having lost its function, becomes an
unmitigated nuisance — an added handicap to the small man in
the struggle for the land.
Lastly, this revolution in the economic basis of society
transformed the cultural life of the community.- When the
clan chiefs came to town, they brought their clan cults with
them; and, when they had sunk their clan differences in com-
mon class interest, the cults were reorganised as state festivals
under their joint control (pp. 123-7) with the object of pro-
viding a divine sanction for the new social order.
Such in general was the process that converted the tribal
settlement, divided by lot into equal shares for all according to
their needs, into the city-state, a town governed by a landed
nobility and surrounded by a poverty-stricken peasantry in
dependent villages. The new unit was the expression of a new
division of labour, agrarian and industrial, which, once estab-
lished, promoted further divisions of labour and thereby
raised human life to new levels of complexity on a slave basis.
Although the process continued in different parts of the
country throughout antiquity, its specific form was deter-
mined in each case, not only by local conditions, which were
of course infinitely varied, but by the general level of society
at the period in question. The later its date, the more pro-
nounced was its class character. This enables us to interpret
what happened at Mantineia in the fourth century (p. 351).
The Spartans, who ordered the dissolution of the city, were
acting, as they always did, in the interests of the big land-
owners. Yet, according to our account of the pSlis, these were
the interests that had brought it into being. Yes, but times
had changed. During the sixth century the development of
commodity production precipitated, in all the advanced city-
states, a further revolution — the overthrow of the landed
aristocracy by the merchant class. In the fourth century, it was
only in backward areas like Arcadia that the landowners were
still in power. In these areas, therefore, the impetus to urbanisation
X
THE FORMATION OF TOWNS
359
did not come from diem; it came from the merchants and
artisans and was carried forward as part of the struggle against
them. Their aim was to arrest the process or, if possible, to
reverse it, as they did for a time at Mantineia.
The same consideration gives us the key to a curious feature
of Greek political terminology. At Athens the term drchoti
denoted both a clan chief (p. 1 14) and one of the annually
elected officers of state. The former usage was the old one; the
latter had grown out of it with the growth of the city-state.
It was the clan chiefs who had constituted, at the inception of
aristocratic rule, the governing body of thecity. At the democratic
revolution these offices were brought under popular control, but
the old title was retained. In other states, younger than Athens,
these annual magistrates were termed dtmiourgol, ‘artisans’. 38 The
new term reflects the shift that had taken place in the balance
of class forces.
4. Phceacia and Pylos
The island of Scheria, where Odysseus was entertained by the
Phaeacians on the last stage of his travels, was identified with
Kerkyra, the modern Corfu. 3 3
'
The inhabitants are described
as emigrants fromHypereia
(not located), from which
they had been driven by
the depredations of their
savage neighbours, the
Kyklopes. 34 They were ex-
pert navigators. On one
occasion they conveyed
Rhadamanthys from Crete to
Euboia and back in a day. 36
FIG. 54. Minoan ship: seal
The Homeric account of them is largely fabulous, but their
service to Rhadamanthys recalls the Minoan thalassocracy, and
38 Th. 5. 47. 9, Hsch. 6nuioupy6y, SIG. 183 etc.
33 Hell. 45, Str. 44, 269, 299, cf. Th. 3. 70. 4.
34 It was variously located in Sicily or Algos (Oi. 6. 4 sch., St. B. ’Apyos)
but these are merely deductions from the myth of the Kyklopes, which
seems to have originated in Anatolia: Roscher LGRM 2. 1688.
36 Oi. 7. 321-4.
X
360 STUDIES IN ANCIENT GREEK SOCIETY
it is worth noting that settlements of Leleges are recorded in
Leukas and Akarnania on the same stretch of coast (p. 170).
What Thucydides says of the towns founded after Minos
had suppressed piracy — all fortified and situated down by the
sea — answers exactly to the position of Scheria as described in
the Odyssey. The town lies on a peninsula, with which it is co-
extensive. It is approached over a narrow isthmus, with a
harbour on either side. 30 Between the harbours is the market-
place, an open space paved with stones, with a shrine of
Poseidon in the middle. 8 ? Here are benches of polished stone
reserved for the king and his counsellors. 38 On the landward
side the approach is barred immediately beyond the market-
place by a wall running across the isthmus. The fields and
pastures, including the royal temenos, which adjoins a grove of
Athena, all lie on the mainland. 88 The timenos is said to be as
far from the town as a man’s shout can be heard. 40 This was
evidently a recognised measure of distance, being mentioned
several times in the Odyssey, and if we may judge from the
Hindu kos it represents about a mile and a half. The kos is a
standard measure in India, and is based on the old rule that the
village territory extends as far from the centre as a man can
make his voice heard. 41
The king’s palace is in the town, and the garden attached to
it contains the spring where the townspeople draw their water. 42
It is, however, a little misleading to describe the leader of this
community as a king. At least he is not a monarch. Alkinoos is
a basileus , but only one of thirteen. 48 These thirteen chiefs
constitute the council of elders ( louU ). They meet in the royal
palace to eat and deliberate, and preside in a body over the
assemblies in the market-place. 44 On the morning after the
stranger’s arrival Alkinoos leads them to their seats in the re-
served enclosure, and meanwhile his herald — the town-crier —
is summoning the people. The business before the meeting
is to make arrangements for entertaining the stranger and
30 Od. 6. 262-5. 87 Od. 6. 266-7.
38 Od. 8. 5-7, cf. II. 18. 503-4; Tritsch AE 83, 87, 99.
30 Od. 6. 259, 291-3. 48 Od. 6. 294, cf. 5. 400, 9. 473, 12. 181.
41 Baden-Powell 12. 43 Od. 7. 1 12-31. 43 Od. 8. 390-1.
44 Od. 7. 136-7, 185-227, 8. 40-5, 4-7.
X THE FORMATION OF TOWNS 361
conveying him home. Alkinoos invites the other chiefs to a feast
at his house, and meanwhile a ship is to be fitted out and
manned by fifty-two young men selected from the towns-
people . 46 After the feast the company is entertained by the
king’s minstrel and then returns to the market-place, where
they witness a programme of sports and dancing. Then gifts
are prepared for the stranger. Each of the chiefs presents him
with a cloak and tunic and a talent of gold . 48 At the end of the
day the chiefs return to the palace for supper, after which
Odysseus reveals his identity and relates his adventures. En-
chanted by his story, Alkinoos proposes, and his colleagues
agree, that in addition to the presents already arranged each of
them shall give a cauldron and tripod, the cost to be defrayed
by the people. 4 ’
In keeping with the aristocratic spirit of Greek epic there
are only incidental allusions to the common people. From
these we gather that Scheria is a community of a single town.
There are no outlying villages. This seems at first inconsistent
with our hypothetical picture, but we must remember that
it has not grown up on its present site. It has been founded
by emigrants in a position favourable for seaborne trade. And
when we look more closely into its constitution, we can detect
signs of an antecedent stage in which it had conformed to type.
The number of the crew which is to take Odysseus home seems
to have been fixed with reference to the chiefs. Each chief
provides four men. This implies that the town is divided into
thirteen wards. And again, when the chiefs agree to present
the stranger with a cauldron and tripod each, we are given to
understand that the levy falls on the ward. In these wards, or
demoi as they would have been called in Attic, we recognise the
separate villages of the original settlement which for the sake
of security and trade have now been concentrated on the one
site.
When Telemachos landed at Pylos, he found the people
gathered on the beach. This was not their normal place of
assembly. Outside the palace of Nestor, which was some way
from the shore, there was a bench of wrought stone, on which
46 Od . 8. 35-6. 48 Od. 8. 390-3.
47 Od. 13. 13-5, cf. 19. 196-8.
X
362 STUDIES IN ANCIENT GREEK SOCIETY
Nestor used to sit as his father had done before him. 48 It may
be inferred that, when the people met in the ordinary way, they
met there. The present occasion is a special one. They are
sacrificing bulls to the sea-god Poseidon, who is also the divine
ancestor of the House of Nestor. On the beach are nine hddrai ,
with 500 men and nine bulls at each. 48 It is clear from the
large numbers involved that these hidrai are not ‘seats’ in the
sense of benches or chairs but separate areas marked out for
the nine groups into which the people are divided. This is
confirmed by other passages in Homer, which show that the
market-place was normally divided in this way. 60 And just as
the people are marshalled on the beach in nine groups, so, as
we learn from the Iliad, Nestor's kingdom comprises nine ter-
ritories. Further, just as each group contributes nine bulls to
the sacrifice, so Nestor led to Troy a contingent of ninety
vessels, ten from each territory. 61 His kingdom was organised
on a tribal basis.
5. Early Athens
We pass on to the greatest of all city-states, whose formation
is described by Thucydides with the acumen of a trained
archaeologist:
From the reign of Kekrops and the earliest kings down to Theseus Attica
was inhabited in several townships, each with its own drchon and its own
prytanefon. Except in time of danger the drcbontes did not meet in council
with the king but administered their affairs independently through their
local councils. Sometimes they even went to war with one another, as when
the Eleusinians supported Eumolpos against Erechtheus. Theseus, however,
a strong and far-sighted king, reconstituted the country by dissolving all
these local councils and authorities and removing them to Athens, where he
set up a single central council and piytancton. He did not interfere with their
property, but merely compelled them to become members of the one city,
which, reinforced from all sides, grew rapidly, and thus enlarged was handed
on to his successors. The Athenians have kept ever since a public festival, the
48 Oi. 3. 406-12. The Trojan market-place was in front of the palace:
11 2. 788-9. The Olympians assembled on the summit of Olympus,
i.e. their acropolis: II. 8. 2-3. See further Tritsch AE 104.
48 Oi. 3. 5-8. 60 11 . 2. 99, 21 1, Oi. 3. 31, 8. 16.
61 ll. 2. 591-602, cf. Oi. 3. 7. sch.; Glotz CG 44. These figures have
been carefully calculated: the total of 4500 (9 X 500) men on the beach
corresponds to the personnel of the contingent (90 X 50); cf. p. 423.
X THE FORMATION OF TOWNS 363
Synoikia, in commemoration of the event. Before that time the city had
consisted simply of the Acropolis with the ground below it to the south.
This is proved by the fact that nearly all the ancient shrines are either on the
south side or on the Acropolis itself. . . . Again, the well now known as
Nine Springs, which, before it was rebuilt under the tyrants, had been an
open spring called Kallirrhoc, was in early times more frequented than any
other because of its accessibility; and the custom still survives of using its
water for pre-nuptial rites and other sacred purposes. Lastly, it is the
ancient use of the Acropolis as a place of residence that explains its present
name: Athenians still call it the City (p 6 lis).&-
The prytaneion — a universal feature of the city-state 63 — was
the town-hall, the building that housed the civic hearth, an
ever-burning fire . 64 When a colony was to be founded, the
emigrants took with them burning faggots from the hearth to
inaugurate the new prytaneion overseas . 66 This was the building
in which distinguished strangers and foreign envoys were
publicly entertained, also citizens who had deserved well of the
community by exploits in battle or at the panhellenic games . 60
Etymologically the prytaneion is the house of the prytanis or
‘president’. This shows that, when the historian speaks of each
town having its own archon and prytaneion, he is referring to the
house in which the principal chief had entertained the others
when they met under his presidency as a council of elders at
the sacred hearth of the community. And so the town-hall
leads back by a long but unbroken line of descent to the first
camp fire.
Further particulars of Theseus are supplied by Aristotle.
He divided the people into three classes — Eupatridai,
Geomoroi, Demiourgoi . 67 The Eupatridai were the families of
the chiefs who enjoyed the hereditary right of serving on the
central city council; and their title, ‘sons of well-born fathers’,
suggests that their consolidation as an aristocratic caste coin-
cided with the official recognition of patrilineal succession.
62 Th. 2. 15.
63 Aristid. Pan. 103. 16 sch., Liv. 41. 20.
64 Pi. N. 11. 1 sch., Paus. 5. 15. 9. In some public oaths Hestia took
precedence over Zeus: SIG. 527. 10, cf. Pi. Leg. 745b, 848d, Paus. 5. 14. 4.
66 EM; TrpvTOVETer.
66 Daremberg-Saglio s.v. Prytaneion. On the irpm-awTov of Olympia see
Gardner 167-9.
67 Arist, fr. 385, Plu. Ties. 25.
X
364 STUDIES IN ANCIENT GREEK SOCIETY
The Geomoroi were the smallholders, who continued to
reside in the country. The Demiourgoi — the artisans — had
already, we may suppose, begun to concentrate in the city. In
the historical era one section of them, the potters, had their
own quarter in the Kerameikos, which was one of the urban
wards or demes, and they had probably been settled there from
very early times . 88 If we could follow the history of these wards,
we should probably find that, though reconstituted from time
to time as the city expanded, they had grown out of the vil-
lages which clustered round the Acropolis when that natural
stronghold had been a pSlis in the original sense of the word.
The substance of what Thucydides and Aristotle say on this
subject was doubtless drawn from oral tradition, and in all
essentials it may be accepted as correct. The only doubts that
arise concern the manner in which the changes were effected
and their attachment to the name of Theseus. The final uni-
fication attributed to him must in the nature of the case have
been preceded by similar movements on a smaller scale. Ac-
cording to Strabo, what he did was to centralise a confederacy
of twelve towns already founded by Kekrops . 80 Athens was one
of the twelve; another was the tetrapolis of N.E. Attica to
which I have referred above (p. 353). From this we see that even
the confederacy of Kekrops was not the first of its kind. In the
same way, the subsequent rise of the Eupatridai must have
kept pace with the decline of the kingship, which we know was
a gradual process. After the Dorian conquest of the Pelopon-
nese the royal office became hereditary in the Medontidai, a
branch of the Neleidai, whom the Dorians had driven from
Pylos. The first limitation on its powers seems to have been
the creation of a separate war chief (polctnarehos ) elected by
and from the Eupatridai . 00 In the middle of the eighth centuty,
while still hereditary in the Medontidai, it was made elective,
with a tenure of ten years, and early in the next century it was
superseded by nine annual magistracies ( archontes ) open to all
members of the Eupatridai. Even then it did not disappear.
The Council continued to meet in the King’s Porch under the
presidency of the drehon basiletis, the ‘king archon’,°i and the
88 Philoch. 72, Mcncd. 3 *=FHG. 4. 449. 00 Str. 397.
00 Paus. 1. 3. 1, Just. 2. 7; Grote 3. 48. Arist, Ath. 57.
X THE FORMATION OF TOWNS 365
Medontidai preserved to the last one remnant of their royal
prerogatives. They owned a piece of ground at the foot of the
Acropolis — their ancient tlmenos .° a
Lastly, there is the problem of Theseus himself. There are
grounds for believing that he came originally from N.E.
Attica and was elevated to the status of national hero in the
latter part of the sixth century (p. 264). It follows that his part
in the unification of Attica must have been invented for him in
that period. In the fifth century he was represented as the
founder of Athenian democracy, who, after forcing the reluctant
gentry to exchange their rustic seats for the amenities of town,
crowned a life of service by laying down his office and leaving
the people to govern themselves. 03 We may be excused from
accepting this gratifying tale, which, as Schefold has recently
argued, 04 was probably invented by Kleisthenes, the founder
of Athenian democracy. The idea that the local chiefs were
reluctant to move corresponds to the conditions of the fifth
century when the Attic yeomen were notorious for their
attachment to their homesteads. 06 The Eupatridai, on the
other hand, had nothing to lose and everything to gain by
residing in the city. They kept their property and increased
their power. They needed no external inducement to pursue
their own interests, and the regime they set up in place of
the kingship became ultimately so intolerable that the people
revolted, threw the grandees out of the country, and
divided their spacious parks among themselves. The most
that can be said for this part of the story is that the
authority of the early kings was undoubtedly limited by a still
vigorous sense of tribal equality, which, even after it had been
shattered, left deep down in the minds of the people a heritage
of democratic ideals which neither rime nor adversity could
efface; and it may well have been the stirring of these ancient
memories in the new democracy of the fifth century that
inspired the tradition as we have it.
02 ZG. 1. 497. 08 PIu. Ties. 25.
04 Schefold 65-7. 06 Th. 2. 16. 2, Ar. ifj. 805-7.
Part Four
THE HEROIC AGE
My wealth is my spear, sword, and shield; with these I
plough and reap and tread wine from the grape, with
these I make my serfs call me lord.
, HYBRIAS
Brothers shall fight and fell each other, and sisters'
sons shall kinship stain.
Voluspa
XI
THE MYCENEAN DYNASTIES
i. The Traditional Chronology
From the fourth century onwards Greek historians reckoned
years in Olympiads, the periods corresponding to the Olympic
Games, which were quadrennial. Local events continued to be
dated by the names of annual magistrates. For earlier times
historians had to base their calculations on the traditional
genealogies. The first attempt to work out a comprehensive
chronology is embodied in the Parian Marble, a long inscrip-
tion dating from 264-263 B.c. Some years later a second
attempt was made by Eratosthenes, whose results do not differ
greatly from the Parian Marble. He assigns the fall of Troy, for
example, to the year 1183 B.c., as compared with 1209 B.c.
Modern archaeology has opened up an entirely new approach.
Minoan objects have been excavated on Egyptian sites, and
Egyptian on Minoan. By this means Greek prehistory has
been synchronised at many points with Egyptian annals, which
in turn have been dated astronomically by the Egyptian
calendar. This method gives promise of exact determinations,
but many difficulties remain.
Archaeology has put an end to the academic scepticism of
many nineteenth-century scholars, who dismissed the heroes
and heroines of Greek legend as wholly unhistorical. It is now
acknowledged that, however encrusted with fabulous accre-
tions, these traditions contain in most cases a kernel of fact.
Indeed, some modem historians have gone to the. other ex-
treme. Bury, for example, accepted such figures as Perseus,
Herakles, Minos, Theseus, and Jason as real persons. He
pointed out that the Greeks themselves believed in their
reality, and that the pedigrees preserved in Homer are remark-
ably consistent. 1 But the Greeks believed no less firmly in the
reality of Hellen, their progenitor, whom they assigned to
1 J. B. Bury in CAH 2 . 478, cf. Myres WWG 340-6.
2
370 STUDIES IN ANCIENT GREEK SOCIETY XI
Table XU 1
THE CHRONOLOGY OF ERATOSTHENES
B.C.
1313. Foundation of Thebes by Kadmos.
1261. Birth of Herakles.
1225. Voyage of the Argonauts.
121 3. War of the Seven against Thebes.
1200. Accession of Agamemnon at Mycenae.
1183. Fall of Troy.
x 176. Achaean settlement at Salamis (Cyprus).
X124. Thessaly occupied by the Thessaloi.
1 104. Dorian invasion of the Peloponnese.
1053. AJolian settlement of Lesbos.
X044. Ionian migration.
1521 b.c., and of Prometheus, the creator of mankind, whose
floruit they fixed at 1600 b.c. Hellen and Prometheus, at
least, are pure myth; yet they differ from the others only in
degree. And the consistency of the Homeric pedigrees is open
to a different interpretation. It is rather a sign of artificiality.
In them a mass of originally independent traditions has been
reduced to a unified system, which involved arbitrary adaptation
and distortion. This is how Nilsson regards them , 2 and his view
is supported by the fact that at several points they are flatly
contradicted by archaeology.
Minos was assigned to the third generation before the
Trojan War — that is, on Eratosthenes' dating, to the genera-
tion of 1260 B.C. He was the Cretan king who cleared the
Carian pirates out of the ^Egean (p. 170). The power of
Knossos had been broken in the fifteenth century — probably by
the Achaeaxis. Bury accordingly accepted him, together with
his date, as an Achaean ruler of Crete . 3 But in that case he
cannot have put down piracy, because in the thirteenth
century, as we learn from Egyptian sources, the TEgean was
thrown into chaos by the tumultuous irruption of various
peoples, including the Achaeans themselves, whose concep-
tion of legitimate seafaring may be judged from what we read
of them in the Iliad and Odyssey . 4 The date of this tradition can
only be preserved by sacrificing its substance. It is much more
2 Nilsson HM 58. 2 Bury 2 , 475. 4 Sec p, 322 n. 90.
XI THE MYCENEAH DYNASTIES 371
-reasonable to accept the substance and let the date go. The
tradition of the Minoan thalassocracy is authentic but it
refers to the period before the fall of Knossos. 8
Similar considerations apply with remarkable uniformity to
all the prehistoric figures for whom there is any recognisable
place in the archaeological background. They are all postdated.
These early generations have been foreshortened by their distance
in time from the chroniclers who formulated the tradition.
That being so, we cannot accept without reserve the reality of
the figures themselves. They must be treated as popular
symbols of remote but impressive events like changes of
dynasty, invasions, wars, and migrations.
2. The Archeological Framework
Shortly before 1600 B.c. there arose at Mycenae a powerful
dynasty whose kings and queens lie buried in the Shaft
Graves. 6 The earliest of these graves shows little sign of Minoan
influence, but in the later ones it is
very pronounced — a wealth of gold
and silver cups and diadems, orna-
mented bronze swords, daggers inlaid
with realistic hunting scenes. Of
particular interest is a silver rhyton
engraved with a battle scene under the
walls of a beleaguered city. The attacking
force wear horse-tail crests, which remind
us of the Carians and Lycians (p. 290).
These kings fortified Mycense and Tiryns
and controlled the country as far as the FIG> 55 * GoM death mask
Corinthian Isthmus, through which they ^ rm ty nast y
maintained contact with the early dynasts of Thebes and
Orchomenos.
About 1 500 B.C. they were succeeded by the Tholos Tomb
6 Hall CGBA 265-6. According to the Parian Marble (n, 19) there
were two kings of the name Minos, one in the fifteenth and the-other in
the thirteenth century b.c., cf. Plu. Thu. 20, D.S. 4, 60. Very likely, the
name was a royal title, like Pharaoh or Cssar.
6 In these remarks on the Mycenean dynasties I have followed Wacc.
XI
372 STUDIES IN ANCIENT GREEK SOCIETY
Dynasty. Tombs of this type have been excavated in Messenia
and Laconia. Under this dynasty the power of Mycenae made
itself felt all over the Peloponnese; intercourse with Thebes
fig. 57 . Embarkation scene: signet from Tiryns
and Orchomcnos became closer; and through them Mycenean
culture penetrated into southern Thessaly.
Some time between 1450 and 1400 B.c. all the cities of
XI THE MYCENEAN DYNASTIES 373
Crete, including Knossos, were destroyed by fire and sword.
Since there is nothing in the later remains to suggest the in-
trusion of aliens, we must suppose that the invaders had
assimilated Minoan culture before they came. 7 The hypothesis
that they were Achseans cannot be proved from the archaeo-
logical side, but on other grounds, as we shall see later, there is
something to be said for it.
After the fall of Knossos Myceme became the political and
fig. 58. Tie Lion Cate of Myrow
cultural centre of the ^gean world. At the beginning of the
fourteenth century there arose a new king who rebuilt the
city. The centre of the citadel was occupied by the palace, sur-
rounded by the dwellings of the court officials and storehouses
for the royal revenues of grain and oil. The city wall was built
of immense blocks of stone with a thickness of ten feet or
more. The main entrance was the famous Lion Gate, sur-
mounted by a slab carved in relief with two rampant lions
7 Wace in CAH 1. 594, Pendiebuty 281.
XI
374 STUDIES IN ANCIENT GREEK SOCIETY
confronting one another on either side of a sacred pillar. Just
inside the gate a stone circle was constructed to enclose the
cemetery of the Shaft Grave Dynasty, and on a ridge beyond,
built perhaps by this same king, was the domed tomb known
as the Treasury of Atreus. The commonpeople of Mycenas
probably lived in hamlets scattered round die citadel. 3 * * * * 8
Later in the same century a new and larger palace was
erected at Tiryns. In this case, apart from the palace, the
citadel was -not inhabited, but it was strongly fortified and used
as a refuge for the people of the town, which lay all round it.
It may be presumed that the rulers of Tiryns owed allegiance
to the Mycenean kings, who controlled direcdy the whole
country as far as Corinth. At Korakou, east of Lechaion, they
had a port, from which their ships traded down the Corinthian
Gulf and across to Thisbe, the road-head leading to Thebes and
Orchomenos. 9
Mycenean objects have been found in abundance all over the
Aegean and far beyond it. In the west they penetrated to
Sicily and Spain. In the east there was close and continuous
intercourse with Troy, Cyprus, Syria, and Egypt. These relations
were not always peaceful, and in the thirteenth and twelfth
centuries the carriers of Mycenean culture seem to have been
marauders and uprooted bands of migrants rather than regular
traders.
Finally, although its Minoan inspiration is at all times ap-
parent, Mycenean culture has several non-Minoan features.
The most distinctive are the ‘megaron’ type of house, the short-
sleeved tunic, the safety-pin, and the use of amber. All these
appear to have come from the north. 10
3. The Traditional Dynasties
How far can the legends of Mycens, Tiryns, Thebes, and
Orchomenos be fitted into this framework? An approach to this
problem has been made in Chapter V, and we may begin by
resuming the conclusions suggested there. It was proposed to
identify Early Cydadic and Helladic with the Carians and
Lelegcs (p. 168), while the culture characterised by ‘Minyan
8 Wacc in CAH 2. 456-8. 9 lb. 2. 457-60. 10 Nilsson HM 72-82.
XI THE MYCENEAN DYNASTIES 375
ware* was assigned to die Pclasgoi (p. 193). The first of these
equations, which seems quite straightforward, calls for no
further comment, but die second is more complicated.
Minyan ware has been found in the Cyclades but not in Crete;
the Pelasgoi, on the odier hand, can be traced in Crete but not
in die Cyclades. If our hypothesis is to hold, this discrepancy
must be explained. It is probable that, like other Cretan
peoples, the Pelasgoi reached Crete from Anatolia, where they
can be traced as far soudi as Tralles in the plain of the Maian-
dros. 11 In that case they may well have split off from the main
body, which passed from die Troad dirough Macedonia into
Thessaly, before die distinctive features of Minyan ware had
been developed. Even so, their absence from the Cyclades
shows that the southward expansion of Minyan ware must
have involved some other factor. Here we may invoke the
Tyroidai and Lapithai. These two stocks are first heard of in
Thessaly, where they have been tentatively identified with the
Dimini culture (p. 197). Both were drawn into the orbit of
Orchomenos, and both expanded into southern Greece. The
Tyroidai can be traced in Corinth, Elis, and Messcnia; the
Lapithai in Attica, Corinth, Elis, Arcadia, Argolis, and also in
the Cyclades. Phorbas and Triopas, who appear in the Argive
pedigrees and again at Rhodes, are Lapith names. 12 Magnes,
whose sons Diktys and Polydektcs settled in Seriphos, was
certainly a Thessalian and possibly a Lapith. 13 For these reasons
it may be conjectured that in southern Greece Minyan ware was
diffused by the Pelasgoi with the assistance of the Tyroidai and
Lapithai, who had absorbed it from them in Thessaly or
Boeotia.
The pedigree of Orchomenos is mere shreds and tatters.
Apart from its Thessalian connections, which have been ex-
amined in Chapter V, die Minyai may be regarded as Min-
oanised Pelasgoi with a palace cult of Demeter, and the
architectural exploits of Trophonios and Agamedes were
doubdess inspired by the Late Mycenean construcdonal works
which have been excavated in the Kopais basin. That is all we
can say.
Poseidon and Libya had two sons, Belos and Agenor. Belos
11 St. B. Nivit). 1® Paus. 2. 16. 1, Hyg. Ast. 2. 14. 12 Apld. x. 9. 6.
XI
376 STUDIES IN ANCIENT GREEK SOCIETY
became king of Egypt. Agenor settled in Phoenicia, and had
four children — Europa, Phoinix, Kilix, and Kadmos. Dis-
guised as a bull, Zeus carried Europa off to Crete, where she
gave birth to Minos. Leaving home in search of her, Kilix
settled in Cilicia, while Kadmos made his way to Rhodes,
Thasos, and eventually Delphi. There, advised by the Oracle,
he abandoned the search and followed a cow to the spot where
it sat down. On this spot he built the city of Thebes. 14
What are we to make of this tradition? It cannot be just
dismissed. The Kadmeioi survived in various parts of Greece
down to the sixth century at least, and were always regarded as
Phoenicians. 18 On the other hand, the Phoenicians have left no
traces in the AEgean earlier than the ninth century. One view is
that the legend of Kadmos rests on a verbal confusion. The
word phoinix means both ‘Phoenician* and ‘redskin*, and
Kadmos, it is suggested, was a Phoenician only in the sense of
being a ‘redskinned' Minoan from Crete. 10 That the Kadmeioi
were in some sense Minoan is clear from the story of Europa and
from their cult of Demeter, which has been examined in Chapter
IV (pp. 123-4). But there is no evidence that the Minoans
were,, or were likely to have been, distinguished as ‘redskins*.
I believe that the clue to this problem lies in the recent exca-
vations at Ugarit (Ras Shamra) near the mouth of the Orontes
in Syria. From very early times this town was an entrepot for
trade between Mesopotamia, Egypt, Anatolia, and Crete.
Many Minoan and Mycenean objects have been found there,
the oldest dating from the seventeenth century. 17 It has even
been suggested by Woolley that Middle Minoan culture may
have been directly indebted to this region for some of its
characteristic features. 18 In the second millennium B.c. no less
than seven different languages are known to have been in use
there, including Babylonian, Hittite, Egyptian,- and Proto-
Phoenician, the parent of Phoenician and Hebrew. 19 It is there-
fore quite possible that the Kadmeioi were Phoenicians who
reached Greece by way of Crete some time in the Middle
14 Apld. 3. 1. x, 3. 4. 1. 1° Hdt. 2. 49, 5. 58 etc.
10 See Nilsson HM 1 3 1. 17 GastcrRS, Schaeffer 3. 18 Woolley 132.
19 Schaeffer 39. On the affinities between the Phoenician and Ugaritic
languages sec Albright PI.
XI -THE MYCENEAN DYNASTIES 377
Minoan period. Indeed, it is more than possible, for the
cuneiform texts of Ugarit record a Phoenician myth in which
the bull god El and the mother goddess Asherat present a close
analogy to Zeus and Europa. 80 Once again we find ancient
tradition confirmed unexpectedly by modem archaeology, and
in this case, as will appear in a later volume, the implications
are very far-reaching.
The Argive pedigree stretches back seventeen or eighteen
generations before die Trojan War. It is the longest we have,
but its contents are disappointing. It shows every sign of
having been arbitrarily reconstructed in the interests of Argos,
whose supremacy over Mycenae and Tiryns dated only from
the Dorian conquest. It is set out in Table XIII in the form
given by Pausanias, who based his account of it on local tradi-
tion. 21 Some of the names mean almost nothing to us, and in
the remarks that follow I shall confine myself to those from
which something positive can be extracted.
At the head of the tree stands Phoroneus, begotten of
Inachos, the stream that flows past the city. Phoroneus is
described as ‘the first man’, who taught the nomads how to
live in towns. 22 He turns up again at Megara as the father of
Kar, the Carian (p. 170). Inis suggests that he stands for the
Early Helladic settlers. In the third and fourth generations
after him we meet the first signs of intruders from the north.
Phorbas and Triopas are Lapith names (p. 375) and Pelasgos
speaks for himself. The Argive acropolis was known as Larisa,
which is an authentic Pelasgian place-name (p. 172); and it was
believed to have been so called after a daughter of this Pelasgos.
There was a tomb in the city reputed to be his, and near it
was a shrine of Demeter Pelasgis. 88 Iasos, a brother of his, is
given by Pausanias as the father of Io; in Hesiod he is replaced
by Peiren, in Aischylus by Inachos. 84 These discrepancies do
not mean much, because Io is a purely ritual figure symbolising
the priestesses of Hera (p. 285). She went, as we have seen, to
Egypt, and her descendant Danaos returned to Argos and
became king in place of Agenor, who abdicated in his favour. 26
a <> Schaeffer 61. aiPaus.2. 15-6, cf. Apld.z. 1 - 4 . 88 Paus.2. 15.5.
88 Paus. 2. 24. l, 2. 22. 1 ; see p. 128.
24 Apld. 2. 1.3, A. Pr. 614-5. 88 Apld. 2. 1. 4.
378 STUDIES IN ANCIENT GREEK SOCIETY
XI
Peirasos
Table XIII
THE ARGIVE PEDIGREE
Inachos
Phoroneus
Niobe
I
Argos
Iasos
I
Io
Epaphos
Lilya
Belos
Aiyptos Danaos
Lynkeus = Hypermestra
J
Abas
Phorbas
_.l
Triopas
I
Agenor
Krotopos
Stheneias
Gelanor
Pelasgos
Larisa
Proitos
Akrisios
I
Danac
Perseus= Mcgapenthes Lysippe=MeIampous Iphianassa=Bias
Andromeda Pelops
J . !
Elektryon Gorgophone SthcneIos=Nikippe Atreus=
[ «=Pcriercs | Klymene
Alkmene Eurystheus |
l
Herald es
I
Agamemnon
Mcnclaos
XI
THE MYCENEAN DYNASTIES
Danaos is an important figure from several points of view.
Like Kadmos, he came from the Levant. Like Kadmos, he
left a settlement in Rhodes. Like Kadmos, he brought
Demeter to Greece. To this we may add that Europa, the
sister of Kadmos, is given as Danaos’ wife. 28 The parallel is
very close, and the Greeks themselves must have been con-
scious of it. This is shown by the appearance of Libya and
Belos in the Argive pedigree, where they do not fit. The line
from lasos to Danaos covers five generations; the line from
Agenor to Gelanor covers only three. Libya and Belos have
been taken over from the Kadmeioi (pp. 375-6). The idea was
prompted by the parallels just noted and the interest of
Argive antiquaries in proving that their city was older than
Thebes. '
As Kadmos came from Phoenicia, so Danaos came from
Egypt. Was this another afterthought? The answer depends on
what we make of Io. As a woman transformed into a cow, she
was identified with Isis, whose sacred animal was the cow. 27 It
would be easy to argue that this equation was no older than
the seventh century. Her son, Epaphos, who became king of
Egypt, resided at either Kanobos or Memphis. 28 Kanobos lies
on one of the Nile mouths — the one that leads past Naukratis
and Sais up to Memphis. Naukratis was a Greek trading
station, founded about 650 B.c. Sais was the seat of the
XXVIth Dynasty, which was then in power. This dynasty was
also associated with Memphis. 29 If this evidence stood alone,
there would be little doubt that the story of Io's journey to
Egypt was invented by the Greeks of Naukratis.
But there was another tradition, inconsistent with this.
When Herodotus visited Chemmis, a town far up the Nile in
the nome of Thebes, he was shown round a temple of Perseus
with a Greek cult attached to it, including an athletic contest —
a practice foreign to the Egyptians. Tire priests assured him
that the contest was founded by Perseus, who was on his way to
Libya in quest of the Gorgon’s head, in commemoration of
the fact that his ancestor Danaos was a native of Chemmis and
set sail from there on his journey to Argos. 30 This tradition
20 Apld. 2.1. 5. 27 Sccp. 28411. 172. 28 A. Pr. 872-8, Apld. 2. 1.4.
29 H. R. Hall in CAH 3. 276, 285. so Hdt. 2. 91.
XI
380 STUDIES IN ANCIENT GREEK SOCIETY
cannot be assigned to the seventh century; for it is very
unlikely that the Greeks of that time, who had only just begun
operations in the Delta, were familiar with the nome of
Thebes.
The same problem is raised by two passages in Homer,
where the Egyptian Thebes is described as the wealthiest city
FIG. 59. Perseus and the Gorgon: Attic vase
in the world. 31 In the seventh century Thebes was of no im-
portance at all; for it was destroyed by the Assyrians in 6 77 B.c.,
and never recovered. The Homeric tradition must refer to it as
it was before its destruction. But throughout the eighth
century and as far back as the twelfth the Greeks had no deal-
ings with Egypt and knew very little about it, as we can see
from die ignorance displayed in the Odyssey . 82 Going back to
the thirteenth century, we find that ./Egean marauders were
harrying the Delta, but, since they were routed there, they are
not likely to have got anywhere near Thebes. 83 We are thus
driven back to the fourteenth century at latest. At that time
Thebes was undoubtedly one of the wealthiest cities in the
world. It was the capital of the XVUIth Dynasty. And this
was the time when the Tholos Tomb Dynasty was reigning at
Mycenae.
31 11 . 9. 381-2, Od. 4. 126-7.
33 Od. 3. 321-2, 4. 355-7; Nilsson HM 136. 33 See p. 401.
XI THE MYCENEAN DYNASTIES 381
This solution of the Homeric problem, which is due to
Lorimer and Nilsson, 3 * solves our problem too. iEgean envoys
can still be seen in the tomb paintings of the XVHIth Dynasty.
HG. 60. Minoans in Egypt: Egyptian painting
If Ionian traders settled at Naukratis in the seventh century,
there is no reason why Mycenean traders should not have
settled at Chemmis in the fifteenth; and,
if the Myceneans had already established
there. a cult of Perseus, that
would have given the Ionians
a precedent for seeking an
Egyptian home for Io. The
myth of Io andDanaos is what,
after a lapse of many centuries,
Greek folk-memory made of
the close relations that had existed between Egypt and Mycenae
in the time of the Tholos Tomb Dynasty.
Danaos was succeeded by Lynkeus, his nephew and son-in-
law, and he in turn by Abas, whose name connects him with
34 Lorimer HUP 153, Nilsson HM 157-8.
FIG. 61. JEgean slip: Egyptian painting
XI
382 STUDIES IN ANCIENT GREEK SOCIETY
the Abantes of Euboia, 35 where there was a local form of the
Io myth (p. 284). Proitos and Akrisios, the two sons of Abas,
quarrelled over the succession. Proitos fled to Lycia, returned
with a Lycian army, and fortified Tiryns. 38 He gave his
daughters to Melampous and Bias of the Tyroidai (p. 196); and
it was he who despatched Bellerophon, another northerner,
to Lycia (p. 164). To Akrisios it was prophesied that he would
be killed by the son born to his daughter Danae. So he shut her
up in a bronze chamber under ground, but Zeus descended
through the roof in a shower of gold, and she gave birth to
Perseus. Mother and child were then cast out to sea in a chest,
which was washed up on the island of Seriphos. There Perseus
grew to manhood and set out for Libya to fetch the Gorgon’s
head. On his way through Palestine he rescued Andromeda
from a sea-monster, and, returning to Seriphos, he proceeded
with her and his mother to Argos. Remembering the prophecy,
his father fled to the Thessalian Larisa. 37 Perseus followed
him and killed him by accident at a game of quoits. Deterred
by this misfortune from claiming his patrimony, he effected
an exchange with Megapenthes, who had succeeded Proitos at
Tiryns. Megapenthes became king of Argos, and Perseus, after
fortifying Mideia and Mycenae, settled at Tiryns. Of his sons,
Elektryon, who succeeded him, begot Alkmene, the mother of
Herakles, and Sthenelos was the father of Eurystheus, who
succeeded Elektryon. At this point some more newcomers
appear on the scene. The mother of Eurystheus is a daughter of
Pelops, and his successor is her brother, Atreus, 38 There seems
to be a dislocation here, caused by some difficulty in com-
bining the Perseid and Pelopid pedigrees. We are told rather
mysteriously that Atreus was ‘sent for’ by Sthenelos, who gave
him Mideia, and after Eurystheus’ death he was again ‘sent
for’ by the people of Mycenae. 39 Atreus’ wife is a daughter of
Katreus and granddaughter of Minos. It was while Menelaos was
away in Crete at the funeral of Katreus that Paris stole his wife. 80
as Apld. 2. 2. 1, Pi. P. 8. 73 sch., Sr. B. ’Apavrfs. 36 Apld. 2. 2. 1.
37 Apld. 2. 4. 4. This point in the saga was probably designed to support
the Argivc claim that the Argirc Larisa was older than the Thessalian:
A.R. 1. 40 sch.
38 Apld. 2. 4. 6. 30 Apld. 2. 4. 6 , Epit. 2. 11. 80 Apld. Epit. 3.3.
XI THE MYCENEAN DYNASTIES 383
There is not much history to be gleaned from all this. The
references to Lycia in the reign of Proitos are remarkably ex-
plicit, but here evidence fails us on the archaeological side.
Danae’s subterranean prison seems to be a faint memory of the
Shaft Graves, confused with the custom of secluding girls at
puberty. If Perseus is the founder of a new line, as he seems to
be, he may be taken to represent vaguely the Tholos Tomb
Dynasty. Herakles is in the main a cult figure (pp. 287-92) and
only one of his exploits calls for mention here. He was sent
by Eurystheus to fetch the Cretan Bull. 41 This beast, which
belonged to Minos, is only another version of the Minotaur,
the bull-headed monster of Knossos (p. 285). In Athenian
tradition the Minotaur was slain
by Theseus, 42 who was placed in
the same -generation as Herakles.
In these two legends we may
recognise a genuine, though faint,
recollection of the fall of Knossos,
and it is significant that this event
is placed immediately before the ac-
cession of Atreus. Who was Atreus,
and where had he come from? This fig. 62. Ariadne, Theseus and
question is bound Up with one of d* Minotaur : gold ornament
the crucial problems of Greek prehistory, which has proved
so perplexing that it has been called ‘the Achaean mystery’. 43
Two final points remain to be cleared up. Seeing that this
pedigree covers the whole Mycenean period, we are surprised to
find no mention of Mycenae before the time of Perseus and only
two allusions to Crete. The first point is explained, as I have
already suggested, by the later supremacy of Argos. We are
led to suspect that, had it not been for the Homeric tradition,
which preserved the memory of Mycenae as the seat of Agamem-
non, that city might have .dropped out altogether. 44 Whether
the kings before Perseus are to be regarded as belonging pro-
perly to Argos or as having been transferred from Mycenae
41 Apld. 2. 5. 7. 42 Apld. Epit. 1. 7-9. 43 Buck GD 7.
44 In the Oresteia, produced just after an alliance had been concluded
with the Argives, Aischylus replaced Mycenae by Argos, but Sophokles
and Euripides restored it.
XI
384 STUDIES IN ANCIENT GREEK SOCIETY
under Argive influence is a question I cannot answer. The
second point is one that concerns both the Argive and Theban
pedigrees. If the worship of Demeter, which Kadmos and
Danaos brought by different routes to Greece, was of Minoan
origin, why is it that there is in the one case only an indirect
connection with Crete and in the other none at all? The ex-
planation lies, I think, in later history. After the Dorian
catastrophe Crete was cut off from Greece, and, when the
Eastern Mediterranean was reopened, the Greeks traded direct
with Egypt and the Levant without touching Crete. The result
was that, when the broken threads were picked up, they re-
integrated the Phoenician and Egyptian traditions of Kadmos
and Perseus in new versions which paid little regard to the
faded glory of Minos.
XU
. THE ACHvEANS
I. Distribution of the Achctans
In the Homeric poems the men who fought under Agamemnon
are described indifferently as Argeioi, Danaoi, or Achaioi. The
Argeioi were properly the people of Argos or Argolis; the
Danaoi were named after Danao s. Under the overlordship of
Mycenae these terms were extended to all those who owed
allegiance to the ruling dynasty of the Argive plain. The third
term seems to have developed in the same way; for in one or
two passages it is used, contrary to the general practice, in a
specific ethnical sense. And this was the usage that survived.
When later writers speak of the Achteans, they always mean,
except where they are consciously following the Homeric
tradition, an actual people inhabiting a definite locality. Our
first task, then, is to identify the Achaeans of historical times.
These were, in the first place, the inhabitants of Achaia.
There were two territories of this name. One was Achaia
Phthiotis in S.E. Thessaly, which for convenience I am going
to call the Thessalian Achaia. The Achteans of this district were
subject to the Thessaloi, who overran Thessaly in the same
period as the Dorians overran the Peloponnese. The other
Achaia was a league of twelve towns strung along the southern
shore of the Gulf of Corinth. Their names were Pellene,
Aigeira, Aigai, Boura, Helike, Aigion, Patrai, Pharai, Tritaia,
Rhypes, Olenos, and Dyme . 1 This is the Peloponnesian
Achaia.
In addition, there were smaller Achaean settlements scattered
all over the Eastern Mediterranean. The islanders of Zakyn-
thos, whom Homer calls Kephallenes , 2 are described by
Thucydides as Achseans from the Peloponnese . 3 These were
doubtless fugitives from the Dorians. In the extreme south of
i Hdt. i. 145, Paus. 7. 6. x. 2 27 . 2, 631.
3 Th. 2. 66. 1. They came from Arcadia: Paus. 8. 24, 3.
Aa
386 STUDIES IN ANCIENT GREEK SOCIETY XII
Laconia, just below the acropolis of Kyparissia, Pausanias saw
the ruins of a town which had belonged- to the Achaioi
Parakyparissroi: 4 ' The- acropolis of Ialysos, one of the three
towns of Rhodes, was known as Achaia, and Achaeans from
here took part in the foundation of Soloi in Cilicia . 6 The
Cilicians (Kilikes) had been known in early times as Hypa-
chaioi, Mixed Achaeans . 6 There was another settlement of
Kilikes near Troy . 7 These claimed kin with their southern
namesakes , 8 and some of them migrated to the southern
Cilicia after the Trojan War.® Another Cilician town, Olbe,
was founded by Aias son of Teukros, whose descendants
reigned there as priest-kings . 10 This Teukros was a native of
Salamis. Driven from home by his father Telamon at the end of
the Trojan War, he sailed to Cyprus, where he landed at
Achaion Akte, the Achaean Shore, and founded the Cyprian
Salamis, which as late as the fourth century was still ruled by
his descendants . 11 Still further afield, the Greek settlement of
Archandrou Polis in the Delta preserved the name of a grandson
of Achaios and a leader of the Achaioi . 12
Returning to the north AEgean, we are told that Skione, on
the Macedonian coast, was founded by Achaeans from the
Peloponnese who were driven ashore there by a storm on their
way home from Troy . 13 At Troy itself the place where the
Greeks had encamped was known as die Achaean Plain. 1 *
Close by were two villages, Killa and Chryse. Killa marked the
grave of Killos, the charioteer of Pelops . 16 Chryse was the
home of Chryses, the priest whose daughter caused so much
trouble in the Iliad . 10 It was founded by emigrants from Crete
called Teukroi. This tradition can be traced back to the eighth
century , 17 but in Attica there was another version. The deme
* Paus. 3. 22. 9. 6 Ath. 360c, Str. 671.
0 Hdt. 7. 91; Kretschmer H, NHA. f Jj. 6. 396-7, 41 5-6, cf. 1. 366.
8 Str. 676. ® Hdt. 7. 91, Str. 668. 10 Str. 672.
11 Str. 682 (cf. Hdt. 7. 90), Iso. 9. 17-8.
12 Hdt. 2. 98. Other settlements in Cyprus which may be identified as
Achaean arc Kourion (Hdt. 5. 113. 1), Lapathos (Str. 682), and Golgoi
(Paus. 8. 5. 2).
18 Th. 4. 120. 1. 14 Str. 596. is Theop. 339, Str. 612-3.
10 II 1. 37-8. 17 Callin. 7=Str. 604.
XII
THE ACH/EANS
387
of Xypete, which lay on the coast opposite Salamis, had
formerly been known as Troon Demos or Troia, 18 and the
story was that a man from this deme named Teukros — not the
son ' of Telamon but an ancestor of his — had founded the
Trojan Chryse. 10 This Attic tradition refers to an expedition
against Troy, previous to the Trojan War, in which Telamon
had taken part. After capturing the city, he married a sister of
Priam. 80 Some of his companions, instead of returning to
Greece, went east and settled in the Caucasus. From them
were descended the Heniochoi and Zygioi, actual Caucasian
peoples that survived throughout antiquity and never forgot
their Achaean origin. 21
These traditions are obviously confused, but that is no
reason for discrediting them; on the contrary it testifies to their
independence. There must have been a genuine affinity between
these scattered Teukroi of Troy, Attica, Salamis, Cilicia,
Cyprus, and Crete. In Cilicia, Cyprus, and the Caucasus they
are directly associated with the Achaean name, and it may be
added that the Odyssey mentions the Achaeans as inhabitants
of Crete. 22
2. Tie Aidkidai
If the Achaeans were in Crete when -the Odyssey was com-
posed, they are likely to have been there before the Dorian
conquest; and in that case it may have been they who intro-
duced Greek speech. It is known that Greek was spoken in
Crete before the Dorians. 23
The followers of Achilles at Troy came from the kingdom of
his father Peleus in the Thessalian Achaia. They are described
as Myrmidons, Achaeans, and Hellenes, with settlements at
Halos, Alope, Trachis, Phthia, and Hellas. 24 The name
Myrmidon was a generic one applied to all the inhabitants of
is Str. 604; Roscher LGRM 5. 123 X. 10 Str. 604. 20 Apld. 2. 6. 4.
21 Str. 416 (cf. 129, 496, FHG. 3. 639), Anus. Marc. 22. 8. 25,
D.H. 1, 89. 4; Kretschmer H 241-3.
22 Od. 19. 175. 23 See p. 399.
24 II. 2. 681-5. not dear how fat the domain of Peleus extended
south of the Spercheios: Str. 431-3; Allen HCS 109-14.
388 STUDIES IN ANCIENT GREEK SOCIETY XII
the kingdom . 85 The . people were divided into five groups,
each with its own chief . 86 These groups correspond to the
five settlements just mentioned. It appears then that the
Myrmidons were a tribal league composed of Achaeans and
Hellenes.
Aiakos, the grandfather of Achilles, was born to Zeus by
Table XIV
THE AIAKIDAI
Phokos
Omytos
Naubolos
Antiphateia
=Krisos
. Asopos
Thebe Aigina=Zeus
Psamathe = Aiakos = Endeis
Eriboia=Telamon=Hesione Peleus= Thetis
Aias Teukros Achilles
Eutysakes Philaios Aias Neoptoi
emos
Strophios=Anaxibia
I
Pylades
Aigina in the island of that name . 27 He had three sons —
Peleus, Telamon, and Phokos. The first two were born to him
by a daughter of Skiron, who was a native of Corinth and a
26 11 . I. 180, 16. 200, 266-9, *8. 69. Possibly the name Myrmidon
belonged originally to the pre-Achaean inhabitants, just as the Danaoi
were properly the pre-Achaean inhabitants of Argos (p. 385). Eurytion,
whom Peleus succeeded by marrying his daughter (see n. 32), was descended
from Myrmidon: Apld. 1. 7. 3, 1, 8. 2.
2 ® 11 . 16. 168-97.
27 Apld. 3. 12. 6, D.S. 4. 72. 1-5, Paus. 2. 5. x-2, 5. 22. 6. Aiakos requires
a fuller treatment than I have given him.
XII
THE ACHjEANS
389
son of Poseidon or Pelops . 88 The mother of Phokos was Psam-
athe, one of the Nereids (sea-nymphs). 2 ® Phokos, the 'seal', was
the eponym of Phokis and the ancestor of Strophios and
Pylades, whose traditional friendship with Agamemnon and
Orestes is a well-known incident in the history of the
Pelopidai. 8 ®
In the lifetime of Aiakos Greece was afflicted with a drought
following a murder committed by Pelops in Arcadia. It was
brought to an end when Aiakos prayed to his father for rain
on Mount Panhellenion in Aigina. In the next generation,
Phokos was killed by his half-brothers, who accordingly were
banished . 31 Peleus went to Phthia, where he married Thetis,
another Nereid, by whom he had Achilles . 82 Neoptolemos, the
son of Achilles, migrated to the highlands of Dodona . 33
Telamon went to Salamis, where he married a granddaughter of
Pelops, by whom he had Aias (Ajax). By Hesione, a sister of
Priam, he had a second son, Teukros, founder of the Cyprian
Salamis . 84
Such is the story of the Aiakidai. There were several
variants, one of which domiciled Aiakos in Thessaly . 38 This
agrees with the Homeric tradition, in which the only son
mentioned is Peleus. 3 ® His connection with Phokos is con-
firmed indirectly by the fact that Aiakidas appears as a personal
name among the Delphic nobility . 37 His ties with Aigina, too,
23 Apld. 3. 12. 6 , Epit. 1. I. In another version Telamon is given as
a son of Aktaios bydauke, daughter of Kychreus (pp. 117-8): Pher. 15. This
points to intermarriage between the Aiakidai and earlier (Minoan?) settlers
on the Attic seaboard.
28 Hes. Th. 1003-4, Pi* 5 * 7 -* 3 * She tried to elude Aiakos by changing
into a seal (E. And. 68 7 sch.)— a totemic metamorphosis, cf. p. 276.
8 ® Paus. 2. 29. 4, E. Or. 33 sch.
31 Apld. 3. 12. 6, Paus. 2. 29-30, D.S. 4. 72. 6-7.
33 H. 18. 85-7, 432-4. It was said that Peleus’ first wife was a daughter
of his predecessor Eurytion (see n. 25): Apld. 3. 13. x. This tradition
was known to Homer: ll. 16. 173-8, It looks as if Thetis was intrusive.
She resisted her wooer in the same way as Psamathe: Apld. 3.13.5; see n. 29.
88 Apld. Epit. 6. 12, Plu. Pyrrb. 1, Prod. Cbr. 1. 3. From him were
descended the kings of the Molossoi: Str. 326.
34 Apld. 3. 12. 7, cf. Pi. I. 6 . 45.
38 St. B. Ata, Serv. ad Verg. A. 4.. 402.
3 ® II. 16. 15 etc. 37 Supp. Epig. Gr. 2. 298. 14-5 etc.
39 ° STUDIES IN ANCIENT GREEK SOCIETY XII
though not mentioned in Homer, must have some foundation
in fact, because the Aiakidai were still flourishing there in the
fifth century. 8 s The most probable conclusion is that the
Aiakidai were an Achajan clan which expanded from Thessaly
into Phokis and down the coast to Salamis and Aigina; That
enables us to explain why, when the pedigrees were finally
systematised, the home of Aiakos was fixed in Aigina rather
than in Thessaly. After her brilliant prehistory, Thessaly
became and remained for centuries a cultural backwater,
whereas Aigina was one of the first states to be drawn into the
'flow of maritime trade when it revived after the Dorian
invasion.
3. The Iottians
We turn next to the Peloponnesian Achaia. A band of
Achaeans from Thessaly, led by Archandros, grandson of
Achaios, or by Pelops, and accompanied by a detachment of
Boiotoi, took possession of Argolis and Laconia, and remained
there till the Dorians drove them out. 88 Under the leadership
of a son of Orestes they then moved to the north coast of the
Peloponnese, from which they expelled the former inhabitants,
the Ionians, and the district was renamed Achaia. 40 The
Ionians fled to Attica, and from there they crossed over to
Anatolia, where they founded the Panionic League of twelve
cities, corresponding to the number which their ancestors had
occupied in the Peloponnese. 41
The historical Ionians were the Greeks of Ionia and Attica,
who spoke closely related dialects. But, as Herodotus remarks,
the Athenians and the Asiatic Ionians not included in the
League were inclined to repudiate the title, 42 which suggests
that its basis was not very secure. This is confirmed by the
circumstances of the migration. The founders of Ionia are
described as a motley crowd of Minyai from Orchomenos,
38 Pi. N. 4. n, 7. 9-10, 0 . 13. 109.
30 Paus. 7. 1. 7 (cf. 2. 6. 5, Hdt. 2. 98), Str. 365.
40 Str. 383-4.
41 Hdt. 1. 145, 8. 73, Str. 365, 383, 385-6. The League was con-
secrated to Poseidon Helikonios (Hdt. 1. 148), which can only refer to
Mount Hclikon (Bccotia), not to Hclike.
42 Hdt. 1. 143. 3.
XII
THE ACHAJANS
391
Kadmeioi from Thebes, Abantes from Euboia, Neleidai from
Attica, Arcadian Pelasgoi, Dorians from Epidauros, and many
others. 48 Such being their composition, the Ionic dialect, as we
know it, cannot have taken shape before the fusion of these
elements in their new home. 44 The Homeric poems point to the
same conclusion. Nowhere do they give any hint of lonians
in the Peloponnese at the time of the Trojan War. The only
lonians mentioned are the Athenian followers of Menestheus. 48
This agrees with the tradition that Ionia was an old name for
Attica, 48 and with Herodotus' statement that the .‘noblest* of
the Ionian colonists were those who had set out from the
Athenian town-hall. 48
This conclusion is admittedly at variance with the tradition
of the three sons of Hellen — Aiolos, Doros, and Xouthos,
fadier of Ion — who were placed at the head of the whole
national pedigree. 48 But none of these figures has any real
roots in the past. They represent the final stage in the system-
atisation of the national traditions — the finishing touch, the
keystone of the arch. In prehistoric times the Greeks had been
scattered, disunited, with no common name and therefore no
consciousness of common origin (p. 348). It was only at the
beginning of the historical era that they developed the national
self-consciousness which the story of Hellen and his sons was
invented to express. The choice of Hellen as first ancestor will
be explained in the course of the present chapter. He is un-
known to Homer, and so are his sons, except Aiolos. Aiolos
was the first to emerge, because the ^Eolic-speaking Greeks of
the Asiatic coast were the first to develop the epic tradition.
Doros, the nominal ancestor of the Dorians, *has no life-story
48 Hdt. x. 146. 2.
44 This conclusion will be re-examined from the standpoint of linguistics
below pp. 518-26.
46 ll. 13. 685, cf. 690, 2. 546-52. 48 Str. 392.
48 Hdt. 1. 146. 2. Herodotus himself implies a close affinity between
die lonians and Achseans: Hdt. 9. 26. 3, He says that before the time of
Ion they had been called Pelasgoi Aigialees (7. 94). I take this to mean,
not that the Pelasgoi were lonians (Kretschmer GDD), which seems
to me quite impossible, but that this part of the Peloponnese had been
occupied previously by Pelasgoi.
48 Apld. 1. 7. 3.
392 'STUDIES IN ANCIENT GREEK SOCIETY XII.
and no recorded offspring; and the Dorian chiefs paid him the
strange compliment of tracing their lineage to Herakles (p. 102).
Similarly, if Achaios had any roots in the past, he would not
have been ignored by Homer, who tells us so much about the
Achseans. As for Ion, he was affiliated through his mother to the
Erechtheidai and worshipped as the father of the eponymous
ancestors of the four Attic-Ionian tribes. When the Neleidai
fled to Attica, the Athenian tribal system was reconstituted in
order to admit them. The myth of Ion marked the formal com-
memoration of this event. As a grandson of Erechtheus, he is
parallel to Boutes, the brother of Erechtheus (p. 186): both are
myths of affiliation or adoption. 49
If there were no Ionians before the Neleidai settled in
Attica, what are we to make of the story that they had been
expelled from the Peloponnese by the Achseans? This point is
met by the simple hypothesis that these Ionians and Achseans
were identical. The Ionian Greeks extended their name re-
trospectively to all of their forefathers who had come from the
Peloponnese. The myth itself hints as much when it represents
Ion and Achaios as brothers, implying that they were closer to
one another than either was to Aiolos or Doros. The same con-
clusion is implicit in the constitution of the Panionic League,
which contained the same number of towns as the Achsean
League that survived in the Peloponnese. 60 It was quite
natural that the Ionians should have reproduced their tradi-
tional dodekdpolis in their new home overseas, but why should
their system have been adopted by the Achaeans who drove
them out? The continuity of organisation argues a continuity
of population. There had never been any Ionians in the
Peloponnese. That was simply the name given in later times
by the Ionians of Ionia to their Achaean ancestors.
4. The Peloponnesian Achaans
Prior to the Dorian invasion, the areas occupied by the
Peloponnesian Achaans were Argolis and Laconia. Apart from
49 The view adopted here regarding Ion and the Ionians has been anticipated
by Meyer GA 3. 397-403.
89 Pib. 2. 417-8.
XII
THE ACHiEANS
393
this tradition, recording the bare fact, they have left no traces
in Argolis, 81 but in Laconia, besides the settlement of Achaioi
Parakyparissioi (p. 386), we find all manner of traditional ties
with Bceotia and Thessaly, and many, if not most, of these
must be put down to the Achseans.
In the first century B.c. the serf population, recently liberated
from Spartan rule, established a confederacy of eighteen
towns, which they called the League of Free Laconians
(Eleutherolakones). Among them were the Achaioi Paraky-
parissioi. There is no need to enumerate all the eighteen towns,
but the following should be noted: Gytheion, Teuthrone,
Akriai, Leuktra, Charadra, Thalamai, Las, Oitylos, Gerenia,
Brasiai, Asopos. 82
At Gytheion there was a tradition that Orestes had been
cured there of his madness. 83 Teuthrone was associated with
Teuthras, another son of Agamemnon; 84 Akriai with Akrias, a
rival of Pelops for the hand of Hippodameia. 88 Leuktra,
Charadra, and Thalamai were believed to have been founded
by Pelops. 88 These were all local traditions going back to the
days when the Pelopidai had been a power in the Peloponnese.
In Iliad IX, anxious to placate Achilles, Agamemnon offers
him seven towns in the south of the Peloponnese — Kardamyle,
Enope, Hire, Pharai, Antheia, Aipeia, and Pedasos. 87 Enope
was identified as Gerenia, one of the Free Laconian towns. 88
Pharai has the same name as a member of the Achaean League.
In Iliad II the seven towns are not included in Agamemnon’s
own domain, but several towns in this part of the Peloponnese,
among them Las and Oitylos, are assigned to his brother
Menelaos. 89 Agamemnon’s domain, as defined in the Iliad,
81 Herodotus says that the pre-Doric dialect of Kynouria had been
Ionic (Hdt. 8. 73, cf. Paus. 2. 37. 3). I take this to be the pre-Achsan
dialect: see p. 523 n. 73.
89 Paus. 3. 21. 6-7. 83 Paus. 3. 22. 1.
s 4 !/. 5. 705 sch. A. 88 Paus. 6. 21. 10. ■
88 Str. 360, cf. Ath. 625c. Epidauros and Letrinoi were said to have
been founded by sons of Pelops: Paus. 2. 26. 2, 6. 22. 8.
82 ll. 9. 149-52. Aipeia is identified by Pausanias (4. 34. 5) with Korone,
named after the Boeotian Koroneia. 88 Paus. 3. 26. 8.
88 ll. 2. 581-6. The Pelopidai had acquired this district by marriage with
the native dynasty of Laconia and Messenia: see p. 430.
394 STUDIES IN ANCIENT GREEK SOCIETY XII
consists, from east to west, of Mycenae, Corinth, Kleonai,
Omeai, Sikyon, Hyperesia, Gonoessa, Pellene, Aigion, Aigialos,
and Helike . 60 This region, if we exclude its south-easterly
extension from the Isthmus to Mycenae, coincides with the
Peloponnesian Achaia. Aigion, Helike, and Hyperesia (the
later Aigeira ) 61 were actually members of the League. Thus,
when the son of Orestes led the Achaeans from Laconia to the
Peloponnesian Achaia, he was seeking refuge in one of his
II. z. 569-77. The rule of Agamemnon was remembered at Sikyon:
Paus. 2. 6. 7.
01 Paus. 7. 26. 1-4.
XII THE ACHiEANS 395
ancestral domains, which had already been occupied by
Achseans in his grandfather's time.
The Achteans whom Pelops led from Thessaly into the
Peloponnese were accompanied, according to tradition, by a
band of Boiotoi (p. 390). These too left their mark. Leuktra,
one of the Free Laconian towns, and founded by Pelops, was a
colony of the other Leuktra in Boeotia, and it had a local cult
of Ino, daughter of Kadmos. 88 Ino was also worshipped at
Brasiai and Thalamai, and the latter was known in later times as
Boiotoi. 83 Another Free Laconian town, Asopos, has the same
name as two rivers, one of them in Boeotia, the other in the
Peloponnesian Achaia. 8 * Gytheion had a cult of the Praxi-
dikai, a local form of the Erinyes. 83 The Praxidikai were also
worshipped at Haliartos in Boeotia, and so far as we know
nowhere else. 88 At Gerenia there was a cult of Asklepios
Trikkaios, brought from the Thessalian Trikka. 87 South of
Teuthrone there were two harbours, one named after Achilles
and the other after Psamatho or Psamathe, the mother of
Phokos. 88 The people of Las were descended from a man of
that name killed by Achilles when he visited Sparta as one of ;
Helen’s suitors. 88 At Kardamyle there was a shrine of the
Nereids, who came ashore there to greet Neoptolemos when
he was in Sparta for his wedding with the daughter of Mene-
laos. 70 These traditions all point to Boeotia or Thessaly, some
of them referring specifically to the Boiotoi, others to the
Achseans.
5. The Origin of the Acheans
Let us return to the north. In the Thessalian Achaia we
found the Achteans federated with the Hellenes under Peleus (pp.
387-8), and now we have seen that they were closely associated
with the Boiotoi. Who were these peoples? The hypothesis I
am going to put forward is that they were branches of a single
stock, which had once inhabited the highlands of Epeiros.
Hellas is the name given in the Catalogue of Ships to one of
88 Str. 360, Paus. 3. 26. 4. 83 Paus. 3. 24. 4, 3. 26. 1, Scr. 360.
84 Paus. 2. 5. 2, 2. 6. 1. 85 Paus. 3. 22. 2. 88 Paus. 9. 33. 3.
87 Str. 360, Paus. 3. 26. 9. 88 Paus. 3. 25. 4.
88 Paus. 3. 24. 10. 78 Paus. 3. 26. 7.
396 STUDIES IN ANCIENT GREEK SOCIETY XII
\
the settlements in the Thessalian Achaia (p. 387). Elsewhere in
the Iliad it is applied generally to the whole country stretching
from Phthia to the southern borders of Boeotia.’ 1 This extended
usage becomes intelligible if we suppose that the Achzeans,
Boiotoi, and Hellenes were virtually the same people.
The Boiotoi, who gave their name to Boeotia, came from
Thessaly. Thucydides says that their occupation of Boeotia
began before the Trojan War and was completed sixty years
THE THESSALIAN ACHAIA Map VIH
later.’ 2 In the Catalogue the whole country is already in their
possession excepting Orchomenos and Aspledon, which are still
ruled by the Minyai (p. 188). They must therefore have moved
south in two stages. The first may be identified with the move-
ment that brought them and the Achseans into the Pelopon-
ncse. The second came when, as Thucydides says, those of
them who had remained in Thessaly were driven south by the
Thessaloi. This may be identified with the ABolian migration —
the movement that established Greek speech in N. W. Anatolia.
’ill. 2. 683, 9. 447, 478, 2. 683 sch. BL, Str. 431-2; seep. 171 n. 89.
’ 2 Th. I. 12. 3. It was the Boiotoi who expelled the Gephyraioi from
Tanagra (p. 123): Hdt. 5. 57.
XII THE ACH^EANS 397
Strabo. says that the main body of these emigrants was drawn
from the Boiotoi. 73
The Boiotoi claimed kinship with the people of Aigina,
which they expressed by saying that Thebe, the eponym
of Thebes, was a sister of Aigina, ancestress of the Aiakidai. 74
Their father was Asopos, a name we have just encountered
among the Peloponnesian Achseans. Boiotos, the eponymous
ancestor, was a son of Itonos, and the national cult was dedi-
cated to Athena Itonia. 76 This cult had come from Itonos in
the Thessalian Achaia (p. 259). In the Iliad Itonos is assigned,
along with Phylake and other settlements, to Protesilaos, who
was a native of Phylake and was still worshipped there in the
' fifth century. 73 So we see that Peleus was not the sole ruler of
the Thessalian Achaia. Protesilaos is not actually described as
his kinsman, but, as a glance at the map will show, the two
domains are so intermingled that they could not have been
ruled without close co-operation. Protesilaos and his followers
may therefore be regarded as a section of the Boiotoi still
domiciled in Thessaly, where they were intimately associated
with the Achseans. 77
In Book XVI of the Iliad, when Patroklos has gone to fight his
last fight, Achilles utters a prayer for his safe return:
O Zeus, Pelasgian Lord of Dodona, who dwellest afar, ruler of wintry
Dodona, the dwelling-place of thy interpreters the Selloi, who have un-
washed feet and sleep on the ground. . . , 78
The reason why at this solemn moment Achilles addresses the
lord of distant Dodona must surely be that, being himself a
descendant of Zeus, he is appealing to the god of his ancestral
73 Str. 402. m Hdt. 5. 80.
76 D.S. 4. 67, Pans. 9. 1. 1, 9. 34. 1, Str. 411.
73 II. 2. 695-701, Pi. I. 1. 58—9, cf. Arr. An. 1, 11. 5. The Achasans
who settled at Skione (p. 386) are described as followers of Protesilaos:
Apld. Epit. 6. 15b. It seems safe to assume that all the Thessalian
chiefs mentioned in the Catalogue, excepting only Euenos, Gouneus, and
Prothoos, were Achsans in the strict sense of the term.
77 That a considerable number of Boiotoi stayed behind in Thessaly
is implied in the tradition that the Thessalian serfs included those Boeotian
inhabitants of Arne who had submitted rather than leave their homes:
FHG. 4. 3 14.
78 11 . z6. 233-5.
398 STUDIES IN ANCIENT GREEK SOCIETY XII
home. And in the same way, when his son Neoptolemos settled
in this region after the war, he was returning to the land of his
fathers. Further, it is agreed that the Selloi or Helloi are no
other than the Hellenes. They are described here as 'interpre-
ters’, that is, priests who expounded the meaning of the signs
sent by the god to those who consulted the oracle.
Aristotle says that the Hellenes came from the country
round Dodona, where they had been known as Graikoi. 78 This
of course explains the name under which they became known
to the Romans, whose first contact with them was naturally
across the Adriatic. One of the Boeotian settlements mentioned
in the Catalogue is Graia, and Aristotle identifies it as the later
Oropos. 80 If we suppose that the Boiotoi had brought this
name with them from Dodona, we are able to understand how
the Hellenes became known there as Graikoi. And it is an easy
supposition to make, because the name Oropos can be explained
in the same way. It has already been noted that Protesilaos was a
native of Phylake in the Thessalian Achaia. There was another
Phylake a few miles south of Dodona, and it lay on a river
called the Oropos. 81
These somewhat tenuous links are clinched by a bond that
endured throughout antiquity. The Boeotians sent a pilgrimage
to Dodona every year, and they enjoyed there a special privilege.
The oracular responses were normally delivered by priestesses,
but the Boeotians had the right to receive theirs through male
interpreters. 82 We recognise the Selloi. The privilege was a
memorial of ancient kinship.
If the Achaeans and Boiotoi had a common Hellenic origin,
they must have played an important role in the diffusion of
Greek speech. Let us see how their migrations appear in the
light of the linguistic data.
In the preceding chapters it has been argued from several
points of view that Greek was first introduced into the Pelopon-
nese by the Neleidai and Lapithai (pp. 165, 183-4, 197, 265-6).
The former settled along the west coast of Messenia, the latter
70 Arise. Mete. l. 14, cf. St. B. rpemed*.
80 II. 2. 498, St. B. , £i 3 pcoir 6 s= J FHG. 2. 41 5.
81 Liv. 45. 26, St. B. 'Wpeowds. Similarly, the Thessalian Arne reappears
in Bceotia: II. 2. 507, Str. 413. Eph. 30=Str. 402.
XII
THE ACH-ffiANS
399
in Argolis, Arcadia, Elis, and the neighbourhood of the
Isthmus .* 8 There is nothing to show what dialect was spoken
by the Neleidai, but it was presumably similar to that of the
Lapithai, of whom something will be said when we come to the
problem of Homeric Greek.
The historical dialects of Argolis, Messenia, and Laconia
were Doric. The people of Elis and Achaia spoke North-West
Greek, akin to Doric, and introduced at the same time. The
speech of Arcadia, however, was neither Doric nor North-
West Greek, but akin to AEolic. Whose dialect was this?
The Doric of Argolis and Laconia contains certain forms
which have been identified as Arcadic. This shows that
Arcadic had once had a wider range, and, since Argolis and
Laconia were the two areas occupied by .the Achteans, there is a
case for associating these Arcadic elements with them. Similar
elements have been found underlying the Doric of Crete,
Rhodes, and Pamphylia. In all these areas the Dorians had
been preceded by die Achaeans. Further, the Greek of Cyprus,
which the Dorians did not reach, resembles Arcadic so closely
as to be virtually the same dialect. 8 * It is clear, therefore, that
this was the speech of the Achaeans — an offshoot of Thessalian
AEolic. When the Dorians broke into Argolis and Laconia, the
Achaean dialect was carried by refugees to Arcadia and Achaia.
The dialect of Boeotia was basically AEolic, overlaid with
North-West Greek. Buck has suggested that the AEolic basis
was the speech of the Minyai, and that the North-West
element was introduced by the Boiotoi . 88 This view cannot be
reconciled with the facts. The AEolic of the Anatolian coast
(Aiolis) differs from that of Thessaly and Boeotia in being
uncontaminated with North-West Greek . 88 It must therefore
have been carried across the ASgean before the intrusion of
88 In Elis die Lapithai ruled over the Epeioi ( II . 2. 620-4, D.S. 4,
69) who were probably Carians: Paus. 5. I. 5, Jo. Ant. 11 =FHG, 4. 546.
Hsch. ’EvSviiIwvcc Kapct. Another Lapith setdement in this region was
Doulichion: II. 2. 625-9, P aus> 5 - *•
84 Buck GD 6-7, Nilsson HM 86-7. f , A
88 Buck GD 3. It is possible, as he suggests, that the Boiotoi got their
name from Mount Boion, but it does not follow that they spoke North-West
Greek.
88 Buck GD 5-6.
XII
400 STUDIES IN ANCIENT GREEK SOCIETY
North-West Greek into Thessaly and Bceotia. But the Boiotoi
were already in Thessaly and Boeotia before the Trojan War,
and moreover the migration to Aiolis was largely recruited
from them — so largely that their new home was sometimes
called Boiotike. 87 It follows that their speech was not North-
West Greek but JEolic. The close relationship for which I
have argued between them and the Achseans is thus confirmed
by the affinity between JEolic and Arcadic.
The name of Zeus is Indo-European (p. 286). For Poseidon
no satisfactory etymology has been found, but he may well be
a parallel form of the old Indo-European rain-god. 88 If it was
the Achasans who brought Zeus to Dodona, they may be
supposed to have come down along the Adriatic coast; and
then, after crossing the Pindos watershed, they followed the
Peneios into the Thessalian plain. There they had been pre-
ceded by the Tyroidai and Lapithai, who had taken the eastern
route down the Axios and along the coast to Petra, where
they established a cult of Poseidon. Dodona and Petra may
thus be said to mark the first intrusion of Indo-European
speech at the two main entrances to Greece.
The extension of the Achasan name as a generic term in the
Mycenean period is now explained by the expansion of the
Achaeans under the suzerainty of the ruling dynasty at Mycenae,
while the common Hellenic origin of the Achaeans and Boiotoi,
who, after absorbing the splendid culture of Mycenae, Thebes,
and Orchomenos, carried it with them to Aiolis and Ionia, the
cradle of Greek epic, enables us to see how the Hellenic name
was destined for an even more glorious future.
6 . The Pelopidai
Looking back over the expansion of the Achaeans, we observe
that the great majority of their settlements are close to the sea.
Not for nothing was Achilles the son of a sea-nymph and
Phokos named after the seal. After reaching Thessaly, they
took to the sea. We may suppose that, like the Tyroidai, they
learnt to sail in the Gulf of Pagasai. The Tyroidai must have
had close relations with them, because one of their branches
87 Str, 402, cf. Th. 3. 2. 3, 7. 57. 5, 8. 100. 3. 88 Cook ZJO 174-5.
XII
THE ACH/EANS
401
was still established at Pherai and Iolkos in the period of the
Trojan War. 80
It is now more than twenty years since Forrer announced
that he had deciphered the names of some Greek princes in the
Hittite documents from Hattusas. Most of his identifications
have been hotly contested, and there is only one of them that I
shall make use of here. Several Hittite kings, beginning with
Mutsil (c. 1350-1320 B.C.), were in communication with the
rulers of a country called Ahhiyava. It is agreed that these are
the Achasans. They were not, however, the Achaeans of the
Greek mainland. The location of Ahhiyava is not yet clear, but
it appears to have been somewhere on the south or west coast of
Anatolia. Their king exchanges presents with Muvatallu, the
son of Mursil (c. 1300 B.C.) and a generation later they are
allied with the king of Assuva (unidentified) in opposition to
the Hittites. In 1240 B.C. their king, Attarisyas, invades
Cyprus. 90 We also hear of princes from Ahhiyava at the
Hittite court, where they learn to drive the horse and chariot. 01
The Achaeans were equally familiar to the Egyptians. In
1288 b.c. Ramses II was routed by the Hittites at Kadesh.
Among the allies of the latter were die Luka (Lycians), Iliunna
(Trojans?), and Kalikisha (Kilikes). 02 Forty years later, in the
reign of Memeptah, Egypt was again threatened by a concerted
onslaught of Libyans from the west together with ‘hordes of
northerners from all lands'. Among these were the Luka,
Shardina, Tursha, and Akaiwasha. The Shardina are either the
people of Sardeis or the ancestors of the Sardinians: they
might even be both. 98 The Tursha are the Tyrsenoi or Tyrr-
henoi, and the Akaiwasha are the Achaeans. Later still, in 1194
B.c,, a similar horde of northerners was defeated by Ramses HI
in the Delta. It may have been stragglers from the receding tide
of this Volkerwanderung that founded Archandrou Polis (p. 386).
It is clear then that the Achaeans were active on the coasts of
Anatolia as far back as the fourteenth century, when they were
80 II. z. 71 1-5. 90 Cavaignac 41-2, 50, 58-9, 86, 92-5.
01 Cavaignac 42. 08 Hall KPS and in CAH 2. 275-6, 281-5.
98 The Sardinians have been traced to the Caucasus (Kretschmer H
225); on the Aegean affinities of the Bronze Age culture of Sardinia see
Childe DEC 242-6.
Bb
402 STUDIES IN ANCIENT GREEK SOCIETY XII
in contact with the Hittite Empire. This is the context in
which we must study the Pelopidai.
Many scholars have held that the Pelopidai belonged to the
same stock as their Achaean followers. This view is quite
plausible in itself, and is supported by one ancient authority.
Autesion, a writer otherwise unknown to us, is reported to
have described Pelops as an Achaean from Olenos . 04 It is of
course very likely that as rulers of the Achaeans they had
become to some extent Achasanised, but there are reasons for
doubting their Achaean origin.
If Pelops, who led the Achaeans from Thessaly into the
Peloponnese, was himself an Achaean, he might be expected to
have left traces in the quarter from which they had come. But
he has not. Before leaving Boeotia he gave his sister Niobe in
marriage to Amphion of Thebes, to whom she bore Chloris,
the wife of Neleus . 85 This is interesting as a sign of ancient ties
between the dynasties of Thebes, Mycenae, and Pylos. After
establishing himself in the Peloponnese he entertained Laios of
the House of Kadmos . 98 At Chaironeia his sceptre was pre-
served as a sacred relic. It had been brought there from Phokis,
and to Phokis from Mycenae by Agamemnon's daughter
Elektra when she married Pylades . 97 That is all. Pelops has
three ties with Boeotia, one of which leads back to Mycenae,
and in Thessaly there is no trace of him at all. Moreover, apart
from Autesion, who is an unknown quantity, all ancient
writers are unanimous in asserting that he was a native of
Anatolia — a Lydian, a Paphlagonian, or a Phrygian . 98 Let,us
hear his biography, which is an instructive example of the way
in which scraps of history were blended with ritual debris into
a typical Greek myth.
His father, Tantalos, a son of Zeus, was born on Mount
Sipylos in Lydia . 99 He had two brothers, Broteas and Daskylos,
94 Pi. O. i. 37 sch.
96 Str. 360 (cf. Nic. Dam. 17), Od. 11. 281-3, Apld. 1. 9. 9.
98 Apld. 3. 5. 5, Ath. 602-3, E. Pi. 1760 sch.
°7 Paus. 9. 40. 1 1-2.
98 Th. 1. 9, Pi. O. 1. 24, B. 7. 53, Hdt. 7. 8y, 7. n. 4, Pi. O. 1. 37
sch., A.R. 2. 790.
99 Paus. 2. 22. 3, Hyg. F. 82, Apld. 3. 5. 6.
XII THE ACH/EANS 403
and a sister Niobe . 1 00 Tantalos used to eat with the gods, and on
one occasion he served them with the flesh of Pelops, then a
boy, whom he had cut up and boiled in a pot. When Zeus dis-
covered the nature of the dish, he directed that it should be put
back into the pot and so restored to life. This was done* and the
child was lifted out by Klotho , 1 01 intact save for a bite out of one
shoulder which had already been taken by Demeter or Thetis.
The missing part was replaced by an ivory splint, with the
result that the Pelopidai were distinguished ever afterwards by
a white birthmark on the shoulder . 108 Tantalos was blasted
with the thunderbolt.
When Pelops grew up, he received from Poseidon a winged
chariot, which could cross the sea without getting wet . 1 03 Setting
out for Greece, he was held up in Lesbos by the death of his
charioteer, named Killos, whom he buried in Lesbos or at
Killa in the Troad . 104 Resuming his voyage, he came to Pisa,
near Olympia, which was then ruled by Oinomaos, a son of
Ares and Harpina . 106 Oinomaos had a beautiful daughter,-
Hippodameia, who was much sought after; but, either because
he had been warned that her son would kill him or because he
was in love with her himself, he was reluctant to let her many.
He compelled every candidate for her hand to compete with
him in a chariot-race. The course was a long one, from Pisa to
the Corinthian Isthmus. The suitor drove off in one chariot
with his prospective bride beside him. He was pursued by her
father in another, and if overtaken he was put to death . 106
Thirteen suitors had been disposed of already in this way, but
Pelops was more fortunate. The girl had fallen in love with
him, and she persuaded her father’s charioteer, whose name
was Myrtilos, to remove one of the linch-pins from his wheels.
The result was that Oinomaos crashed, and was either killed in
the fall or speared by Pelops . 107
Meanwhile Myrtilos had become enamoured of the bride.
100 p aus , 3. 22. 4, A. R. 2. 358 sch. 101 Pi. O. x. 23-51.
108 Pi. 0 . 1. 37 sch., Hyg. JF. 83. 108 Pi. 0 . 1. 75-8, 87, Apld. Epit. 2. 3.
104 Theop. 339, Str. 613.
106 Pi. O. 1. 65-88, Paus. 5. 22. 6, A.R. X..752 sch., cf. Paus. 6. 21. 8.
106 Apld. Epit. 2. 4, D.S. 4. 73.
107 Pi, O. 1. 127 sch., Paus. 6. 21. 7, Apld. Epit. 2. 6-7.
404 STUDIES IN ANCIENT GREEK SOCIETY XII
While they were crossing the JEgean (it is not dear what had
brought them there) Pelops left the chariot to fetch a drink, and
during his absence Myrtilos attempted to rape her. Pelops
came back and pitched him into the sea. 108 Returning to Greece
after these adventures, he conquered the Peloponnese (we are
not told how) and named it after himself. Hitherto it had
been called Apis or Pelasgiotis. 108 He succeeded his father-in-
law at Pisa, and had many sons, of whom Atreus and Thyestes,
after residing for a while at Makistos (Triphylia), removed to
Mycenae and Tiryns. 110 His bones were preserved in a te'menos
dedicated to him at Olympia. 111 As for his sister Niobe, whom
he gave in marriage to Amphion, she bore several children of
whom she was so proud that she declared herself happier than
Leto, whereupon all her children save Chloris were slain by
Apollo and Artemis. She returned home to Sipylos stricken
with grief, and was turned to stone. 112
The boiling of Pelops is a myth of initiation. 118 Klotho is
already familiar to us as a goddess of birth (pp. 3 34-6); here she
is a goddess of rebirth. The race for Hippodameia is based on
the svayamvara or pre-nuptial contest — a matriarchal develop-
ment of the initiatory ordeal imposed on youths before mar-
riage. 114 What concerns us at present is not this ritual nucleus
but the residue of historical fact.
If Pelops gave his sister away in Bceotia, he cannot have come
straight from Sipylos to Pisa. This discrepancy shows that we
have to do with two distinct traditions. The first, which men-
tions Thessaly but not Sipylos or Pisa, is the Achaean version;
the second, which ignores Thessaly and the Achaeans, belongs
to Pisa.
If Pelops conquered the Peloponnese, it is strange that he
chose Pisa as his capital rather than Mycenae or some other
town of known eminence at this period. Pisa was of no im-
portance at any period, except that for a time it had controlled
the Olympic Games, which did not acquire their panhellenic
character till the eighth century. If Pisa had ever been the seat of
108 Apld. Epit. 2. 8-9, ll. 2. 104 sch. A.
100 Apld. Epit. z. 9. 11 » E. Or. 5 sch. 111 Paus. 6. 22. 1.
112 11 . 24. 602-17, D.S. 4. 74, Apld. 3. 5. 6.
113 See my AA 113-8. 114 Briffault 2. 199-208.
XII
THE ACHJEANS
405
the Pelopidai, it might be expected to figure in the Iliad as one
of Agamemnon's domains, but it is not mentioned. It is
difficult to avoid the suspicion that the connection with Pisa is
an accretion to the myth.
Shortly after the Dorian invasion a band of Achaeans under
one Agorios migrated from Helike in the Peloponnesian
Achaia and settled in Elis. Agorios was a great-grandson of
Orestes . 118 This gives the clue. The cult of Pelops was brought
to Elis at this late date by a branch of the Pelopidai, and there
it found a home at Olympia with the goodwill of the festival
authorities, who had an obvious interest in appropriating so
illustrious a tradition.
Other signs point in the same direction. Hippodameia was
buried at Olympia, but her bones had been brought there from
Mideia, whither she had retired after quarrelling with her
husband . 116 Oinomaos hardly exists except in relation to
Pelops. He had no known predecessor at Pisa, and why did he
fix the goal of the race at the Corinthian Isthmus? The reason
seems to be that he was himself an importation from that
quarter; for his mother’s father was the river Asopos . 111 Helike,
Mideia, Corinth, and the Asopos all lie within the territory
which is assigned in Homer to the Pelopidai of Mycenas. On
these grounds I believe that it was here, in the north-east
corner of the Peloponnese, that the myth of Pelops was first
planted on Greek soil. But it does not follow that Pelops him-
self had once reigned at Mycenae. On the contrary, there are
signs that he never reached Greece at all. There was a tradition
that his bones were fashioned into the image of the Trojan
Athena . 118 His charioteer died before leaving Anatolia, and
this tradition suggests that he did the same. It is supported,
moreover, by another, which says, in striking contrast to the
Olympian version, that Oinomaos was king of Lesbos . 119
The chariot-race, like the boiling, was simply an ancestral
116 Paus. 5. 4. 3. 116 Paus. 5. zo. 7, 6. 20. 7.
11 Paus. 5. 22. 6, 6 . 21. 8. Pelops’ chariot was preserved at Phleious:
Paus. 2. 14. 4.
118 Dion. Rhod. 5= Clem. Pr. 4. 14, U. 4. 92 sch., Tz. ad. Lyc. 53, 9x1,.
Paus. 5. 13. 4-5.
119 E. Or. 990 sch.: for Nfitroav (Dindorf 2. 250. 5) read 'lo-crav (Str. 60).
406 STUDIES IN ANCIENT GREEK SOCIETY XII
legend which the Pelopidai brought with them from Anatolia.
Sipylos is the mountain overlooking the Hermos valley
between Sardeis and the sea. Here Niobe ‘all tears' was turned
to stone. Here was a rock known as the Throne of Pelops. Here
was an ancient sanctuary of the. Mother of the Gods, built by
his brother Broteas. Not far from here, in one of the tributary
valleys, was the town of Thyateira, originally called Pelopeia. 120 -
The Hermos was the main avenue by which Hittite culture
reached the JE gean. The figure of Niobe, the Throne of
Pelops, the shrine built by Broteas — all these refer to the
Hittite monuments still visible on the heights of Sipylos. Not
only so, but the third brother, Daskylos, is a namesake of the
father of Gyges, the first of the Mermnadai (p. 174); and the
charioteer Myrtilos is a namesake of Myrsilos, the last of the
Lydian Herakleidai, and also of Mursil, the Hittite king who
recorded at Hattusas his dealings with the princes of Ahhiyava.
There was evidently real substance in the tradition that Pelops
was a Lydian.
He was also a Paphlagonian and a Phrygian. Paphlagonia lay
immediately to the north of Hattusas, and it was the country of
the Leukosyroi, who have been identified as Hittites. The
Phrygians were an Indo-European-speaking people, akin to the
Thracians, who crossed the Hellespont and overran the Hittite
Empire. Like the Achaean conquerors of Knossos, they fell
under the spell of the older culture. The Phrygian Kybele was
the Hittite mother-goddess in a new form, and many extant
Phrygian monuments, if not actually of Hittite workmanship,
are at least inspired by Hittite originals. Among them are the
120 Paus. 5. 13. 7, 3. 22. 4, St. B. ©u&ropot. There are many other
Anatolian connections. The hordax dance, associated at Olympia with
Artemis, came from Sipylos: Paus. 6. 22. 1. The shrine of Artemis Mouny-
chia at Pygela was reputed to have been founded by Agamemnon: Str. 639.
Among the previous suitors of Hippodameia were Mermnes, Hippothoos,
Alkathoos, and Pelops of Lokris Opountia: Paus. 6. 21. 10, Pi. O. 1. 127
sch. Mermnes is the eponym of the Mermnadai (p. 174). Hippothoos was a
grandson of Teutamos: II. 2. 840-2; see p. 260. Alkathoos is given by
Hesiod as a son of Porthaon from Pleuron (Paus. 6. 21. 10) but in Homer he
is a brother-in-law of Aincias and his wife is Hippodameia (II. 13. 428-9):
the early inhabitants of Pleuron were Leleges from N.W. Anatolia (p. 427).
On the Trojan connections of Lokris see p. 259.
XII
THE ACH^ANS
4 °7
lion tombs of Ayazzin and Dimerli. 121 The entrance to both
these tombs is surmounted by a stone slab on which is carved a
pair of rampant lions facing one another, with an upright
column between them. As Garstang points out, this was a
characteristic Hittite conception. We have already met it at
Mycenaj (p. 373).
The Greeks had no direct knowledge of the Hittites as dis-
tinct from the later peoples that inherited their culture. Hence,
when they described Pelops as a Lydian, Paphlagonian, or
Phrygian, that was as near as they could have got to saying he
was a Hittite.
The collapse of the Hittite Empire must have been complete
by 1200 B.c., and Cavaignac assigns the Phrygian invasion to
the same period. 122 If this is right, it seems to follow that Troy
was already occupied by the invaders at the time of the
Trojan War. In the post-Homeric tradition Hecuba is a
Pelasgian from Thrace (p. 260) but in Homer her father is
Dymas, a king of the Phrygians. 123 Priam himself is not a
Phrygian, but the Trojan pedigree is so confused that this
discrepancy does not count for much; and on one occasion he
recalls a campaign in the Sangarios valley which he had fought
in alliance with the Phrygians against the Amazons. 124 This
seems to indicate that they were already in contact with the
Hittites before the Trojan War; and it is worth noting, though
hard to explain, that in this campaign their king was Otreus,
which is the yEolic form of Atreus. 125
One feature inherited by Kybele from the Hittite goddess
was her lion-drawn chariot. 126 The Hittites were famous for
their chariots, and perhaps the same origin may be postulated
for die vehicle that figures in the career of Pelops. In Cappa-
docia the goddess became known as Ma, with her centre at
Komana, an important town in Hittite days, and here, right
down to Roman times, there survived a hieratic clan called the
121 Garstang 16, cf. 85. Perhaps he goes too far in deriving the tombs
directly from a Hittite original; and of course they are much later in
date than the Lion Gate at Mycenae, but that does not conflict with the
hypothesis of a common Anatolian prototype.
122 Cavaignac 152. 123 II. 16. 718. 124 II. 3. 185-9.
125 U. 3. 186, Horn. H. 5. in, 146, EM. ’Orpsis.
126 Garstang 1x4, cf. A. J. Evans RN 33-7.
408 STUDIES IN ANCIENT GREEK SOCIETY XII
Agamemnoneion Genos or Orestiadai, who claimed to have in
their keeping the image of Artemis brought from Tauris by
Orestes. 127
Who then were the Pelopidai? The Greek tradition points
to the Hittites. But in the annals of Hattusas there is no hint
that Hittite rule ever extended to Greece proper. Perhaps they
were an Anatolian branch of the Achseans who had assimilated
Hittite culture. In the present state of our knowledge it
would be unwise to press for a more definite conclusion.
Their pedigree, as given in Homer, covers four generations:
Table XV .
THE PELOPIDAI
Tantalos
Agorios Gras
(i) Pelops; (2) Atreus and Thyestes; (3) Agamemnon and
Menelaos, sons of Atreus, and Aigisthos, son of Thyestes; (4)
Orestes, son of Agamemnon. The fifth generation, repre-
sented by Tisamenos and Penthilos, the sons of Orestes, is
ignored, because it is subsequent to the period to which the
poems refer. This pedigree was presumably derived from the
127 d.C. 36. 13, St r. 353, CIG. 4769.
THE ACHjEANS
XII
409
family tradition of the Penthilidai, who were kings of Lesbos
at the time when the poems were composed. 128
The mainland Greeks preserved a different version. In
Hesiod the father of Agamemnon and Menelaos is not Atreus
but a son of his named Pleisthenes. 129 Pleisthenes is a puzzle.
No motive can be discerned either for inserting him or for
removing him. In either case he proves that the number of
generations was not always remembered correctly.
Tisamenos and Penthilos were contemporary with the
Dorian conquest. Expelled from Laconia, Tisamenos fled to
Helike, where he died and was buried. 130 His son, Kometes,
had already emigrated to Anatolia. 131 Penthilos, who founded
the AEolic colony in Lesbos, left two sons in the Peloponnese,
one of whom, Damasias, was the father of Agorios. 132 It was
this Agorios who led a band of Achasans into Elis (p. 405),
where, according to the local tradition, he was welcomed by a
chief called Oxylos. Here we can lay our finger on another-
weak spot in the traditional chronology.
Eratosthenes fixed the average length of a generation at
forty years. This estimate goes at least as far back as Thucy-
dides, who dated the Dorian invasion eighty years after the
fall of Troy, corresponding to the two generations between
Agamemnon and his grandsons, whom the Dorians expelled. 133
As Bum has pointed out, it is far too long. 134 The difficulty is
not met, however, by his expedient, which is simply to scale
down the dates of Eratosthenes. That is to assume that the
number of generations is correct, but it seems to me their
number is no more reliable than their length. Pleisthenes is
one case in point, and Oxylos is another. In the Elian tradition
he is a grandson of Thoas and a native of Aitolia. Acting in
concert with the Dorians, he occupied Elis with a band of
Aitoloi. 138 This agrees with the Homeric Catalogue, where
128 Arist. Pol. 1311b. 19, Str. 40Z, 447, 582, Patis. 3. 2. 1.
129 Hes. fr. 98, cf. Stes. 15, A. A. 1568.
iso Paus. 2. 18. 8, 2. 38. i, 7. :. 7-8. 181 Paus. 7. 6. 2.
132 Paus. 3. 2. I, 5. 4. 3. 133 Th. 1. 12. 3, cf. Hdt. 1. 7. 4, 2. 142. 2.
134 Bum DEGH, cf. Chadwick GL 1. 193, 198. Possibly the source
of the error was the late marriages customary at Athens in the fifth century.
iss Paus. 5. 3. 6. This was the migration that brought the North-West
dialect to Elis.
410 STUDIES IN ANCIENT GREEK SOCIETY XII
Thoas is mentioned as a chief of the Aitoloi at the siege of
Troy . 180 It also agrees with the assumption that the Dorian
invasion took place two generations after the war. But it does
not agree with the pedigree of Agorios, who is assigned to the
fourth generation after the war. They might of course have been
contemporary in spite .of this discrepancy, because the length
of a generation varies; but that only shakes our faith in the
traditional chronology, which made the pedigrees commen-
surable by postulating the same number of generations for
different families in a given period. It is much more probable
that one or other of these pedigrees has been tampered with.
And which is the more reliable — the one that obeys the
artificially coherent system of Eratosthenes, or the local variant
that asserts its independence?
In view of these considerations there is no reason to regard
Pelops as more than a symbol of the Anatolian origin of the
dynasty. Once established at Mycenae, these kings had no
interest in preserving the full record of their Asiatic forbears.
No doubt their own names were preserved, probably in
writing, so long as they remained in power; but, when the
dynasty was overthrown, their tradition split, like themselves,
into two branches. The Penthilidai of Lesbos retained their
royal status, but, as their name implies, the functional value of
their traditions was limited to proving descent from a son of
Orestes, from whom their status was inherited. The earlier
generations, remote in place and time, tended to drift into the
realm of myth. Meanwhile the other branch of the family,
headed by Agorios, survived in the Peloponnese, but only after
a social upheaval in which the family tradition must have
suffered almost as much as the family fortune; and what re-
mained of it when they settled at Elis was modified in the
interests of the Olympic priesthood.
The Argives agreed with Homer in placing Agamemnon's
tomb at Mycenae, but the Spartans showed his tomb at Amyklai . 137
is® 11 z. 638.
1 3 ? Paus. 2. 16. 6, 3. 19. 6. Amyklai remained in Achaan hands for
some time after the Dorian conquest: Paus. 3. 2. 6. The tradition of
Agamemnon's domicile at Sparta seems to have been known to the poets of
the Odyssey: Alien HCS 66-9.
XII
THE ACHjEANS
411
If, as I believe, Agamemnon is not to be regarded as a
definite historical individual, we are absolved from the
necessity of choosing between these alternatives. It is enough
for us to say that, when the Spartan branch of the dynasty,
represented by Menelaos, established themselves on the
Eurotas, they took their ancestral cult with them and main-
tained it in their new cemetery, which, deserted by them, was
remembered long afterwards by the Achaeans as the burial place
of their ancient kings, of whom the greatest was Agamemnon.
The same sort of thing happened all over Laconia. The
Pelopid tradition survived, but in decadent forms, localised,
incoherent, representing all that the downtrodden serfs were
able or willing to recall of the days of their greatness.
The upshot of my argument is to suggest that, after leading
an army of Achaeans, some of them perhaps from Anatolia, in
a devastating onslaught on the centre of the Minoan Empire,
the Pelopidai were invited, or invited themselves, to Mycenae,
where they built the Lion Gate and the Treasury of Atreus.
As the problem stands, this seems to me the most probable
solution, but of course it is far from certain. The Achaean
mystery is not so deep as it was, but we have not got to the
bottom of it.
xm
THE CLASH OF CULTURES
i. The Social Character of the Achteatis
After all that has been said the actual history of the Greek
matriarchate still remains to be written. It is proved to have
had all the salient features of mother-right revealed by our
general study of the subject, but we have not been able to dis-
play it in detail as a going concern or to identify its phases of
growth and decay. That must wait until we can read the
Minoan script. In the meantime we can say this. Between the
first and last' of the nine Minoan epochs there lies a process of
continuous change, involving the development of property at
the expense of kinship rights and a gradual shifting of the
balance in the social relations of the sexes. But this shift
took place within the matriarchal framework, which main-
tained itself, with deepening contradictions, down to the
period we have now reached. If Greece had been less accessible
to the outer world and more easily centralised, the old system
might have lasted as long as it did in Egypt; it might have
absorbed the successive shocks of barbarian incursions, as
Egypt did the Hyksos; but in the Achaean age this pressure from
outside, combined with its internal contradictions, precipitated
a crisis, which produced Greek civilisation as we know it — a
class society founded on private property and animated in its
early stages by a conscious struggle to transform or suppress
the institutions and traditions of the matriarchal past.
The general characteristics of Achasan society as portrayed in
Homer have been analysed by Chadwick, who has pointed to
many illuminating analogies between this and other ‘heroic’
ages . 1 Rude but vigorous invaders subjugate and assimilate
a superior culture, thereby bringing about an economic and
social upheaval marked by the accumulation of wealth in the
1 Chadivick HA, GL.
XIII THE CLASH OF CULTURES 413
hands of an energetic military caste, which, torn by internecine
conflicts of succession and inheritance, breaks loose from its
tribal bonds into a career of violent, self-assertive individualism
— a career as brief as it is brilliant, because their gains have
been won by the sword and not by any development of the
productive forces.
The wealth of these Achasan chiefs consists first and fore-
most in their flocks and herds. ‘Why should I fight the
Trojans?' Achilles shouts at Agamemnon. ‘They haven't raided
my catcle.’ 2 To their own livestock must be added tributes in
kind from the peoples they have conquered. In offering
Achilles the seven Messcnian townlands Agamemnon assures
him that the inhabitants will ‘honour him with gifts like a
god', 3 meaning that they will pay him a percentage on their
holdings. Female captives are valued for their skill at the loom,
and they are priced in terms of cattle. 4 For the rest, these
rapacious adventurers covet gold and silver vessels, bronze
tripods, cauldrons, goblets, any objet i ' art of Minoan crafts-
manship they can lay hands on. Their scale of values may be
seen in the prizes awarded at their games. Chariot-race: first
prize, a skilled woman and a tripod, capacity 22 measures;
second prize, a six-year-old mare in foal; third prize, a caul-
dron, capacity 4 measures, brand-new; fourth prize, 2 gold
talents; fifth prize, a cup. 6 Boxing: a six-year-old mule, not
broken in, and a two-handled cup. 6 Wrestling: a tripod
valued at 12 oxen and a woman valued at 4 oxen. 7 Foot-race:
first prize, a silver mixing-bowl from Sidon, capacity 6
measures; second prize, an ox; third prize, half a gold talent. 8
Their moral values, their personal ideals, and their attitude
to the common-people, are mirrored in the stories told to them
by their minstrels about the gods. Zeus dwells on the cloud-
capped peak of Olympus. 0 In the beginning, as cloud-
gatherer and thunderer, he had dwelt alone, the other gods
residing elsewhere — Hera in Argos, Aphrodite in Paphos,
Athena in the House of Erechtheus; but now they have been
gathered together in a single celestial stronghold — Zeus in the
2 ll. 1. 154. 3 ll. 9. 154-5. 4 2 /. 23. 703, 885, Oi. 1. 431.
6 ll . 23. 262-70. 6 ll . 23. 653-6. 7 ll . 23. 702-5.
8 ll . 23. 741-51. 0 Nilsson HM 267.
414 STUDIES IN ANCIENT GREEK SOCIETY XIII
central palace, the others in the surrounding mansions built
for them by Hephaistos. The supremacy of Zeus is acknow-
ledged, but it is often challenged, especially by his wife. He
summons his peers to councils, at which the fate of mankind is
decided, and entertains them to meat, wine, and music. These
gods are selfish, unscrupulous, covetous, intensely alive to all
delights of the senses. In one thing only are they divided from
their worshippers — they can never die; and of this privilege
they are passionately jealous. Man must not aspire above his
mortal station, or he will be blasted by the thunderbolt. As the
common people are to their chiefs, so is man to the gods . 10
The Achaean Olympus is the mirror of social reality.
Stripped of their heroic glamour, there is little in these men
to command our admiration except their boundless vitality.
Listen to the disguised Odysseus at pains to make a good im-
pression:
My father was a rich man of Crete. He had many sons by his lawful wife,
but my own mother was a concubine, though he loved me none the less for
that. Kastor son of Hylax was his name, a man of wealth and fortune, the
happy father of fine sons. When he died, they cast lots for the inheritance
among themselves — all they gave me was a meagre portion and a house.
However, I married a woman of good family with many shares of land —
thanks to my manly spirit, for I was no weakling and not afraid of war. It’s
all gone now — I’ve had my full share of trouble since those days — you
must judge the crop from the stubble. Ah, what a bold and battling heart
was mine, leading a band of stalwarts in an ambuscade, with never a thought
of death in my head — I was always the first to pounce on the hindmost
when the enemy took to their heels and to strip him of his arms. But hus-
bandry and housekeeping and bringing up a family were things I could
never abide. What I enjoyed was the sea and war — ships, spears, arrows —
they make other men shudder, but I loved them. That was the nature God
gave me — every man to his taste. Already, before the A chains sailed to
Troy, I had led nine raids on foreign shores and got plenty of loot, gifts of
honour as well as my share of the lots, so my affairs prospered and I was
respected and feared in Crete. But then Zeus ordained that accursed expedi-
tion that has laid so many low, and they chose me and Idomcneus to lead
our ships to Troy. There was no help for it. The stern voice of the people
called me, and I had to go. 11
Phoinix, when we meet him in the Iliad, is a grave, god-
fearing old man, but he too has had a stormy past:
10 Nilsson HGR 158-9. 11 Od. 14. 199-239.
XIII THE CLASH OF CULTURES 415
I left home after a quarrel with my father, Ormcnos son of Amyntor. He
had fallen in love with a concubine, which was a disgrace to my mother, his
lawful wife. She begged me to have the girl first so as to make her dislike the
old man, and I did as she asked. When he found out, he cursed me — he
prayed to the Erinyes that no child of mine might ever sit on my knees, and
that prayer has been fulfilled. I was going to kill him, but some god held
me back and warned me of die accusing voice of die people if I should be
named a parricide. I couldn’t bear die thought of living under one roof with
that raging father, but my cousins and kinsmen came and entreated me and
tried to keep me by force. Nine nights they spent in the house, slaughtering
sheep and oxen, roasting pigs at the fire, and swilling the old man's wine.
They kept watch on me by turns, with a fire always burning, one in the
courtyard and another in the porch outside the bedroom door; but on the
tenth night I broke through die door, sealed the courtyard wall unseen, and
fled. I made my way through Hellas to Pclcus in Phthia, who welcomed me
and loved me as though I were his long-awaited one and only son; he made
me rich too and gave me a fine people to rule — I became king of the Dolopes
on the distant frontiers of Phthia. 12
In view of passages like these, it is a little disconcerting to read
in the polite pages of Homeric scholarship that the Achseans
were ‘a gentle and generous race’ with a pure and tender
conception of conjugal affection '. 13
The ‘cousins and kinsmen’ who consumed so much meat
and liquor were evidently present in numbers for transcending
the limits of a family. We have here caught an Achaean in the
very act of breaking away from the restraints imposed by his
clan; and after making good his escape he attaches himself to a
stranger in a purely individual relationship — the bond of
personal allegiance between a vassal and his lord. The lawful
wife’s part is equally interesting. In enjoying as many women
as sword could win and cattle buy, Ormenos may have been
true to the custom of his northern ancestors; but his wife
objected, and, like Clytemnestra, she hit back. These new-
comers married into the native nobility, who had their own
' ideas about the dignity of womanhood. That is why so many of
these ‘heroic’ tales turn on quarrels about wives and concubines.
Helen herself, the fairest or them all, had chosen her husband
for herself from tire Acha:ans competing for her hand; and,
having chosen freely in the first instance, she was free to
change her mind. In this case it was the husband that objected,
12 II. 9. 447-84. 13 Jcbb 53-4.
416 STUDIES IN ANCIENT GREEK SOCIETY XIII
and the Achsans rallied to his side. It took more than Helen’s
face to launch the thousand ships. Paris stole goods as well. 1 *
The wealth went with the woman. The fights about fair
ladies were fights about hard cash.
2. The Homeric Treatment of the Matriarchate
The Homeric poets were well aware that the world in which
their heroes had fought their way to glory was very different
from their own, and there were several distinctions on which
they were careful to insist. In their day, bronze had been super-
seded by iron, and the Dorians were supreme in the Pelopon-
nese; in the poems iron is a rarity and the Dorians are ignored.
In these matters they delineated consciously an idealised
picture of the past. In regard to the growth of property and the
consolidation of the patriarchal family the picture is much more
subjective, and hence lacking in clarity. They knew, however,
that the status of women had undergone a change, and if we
scrutinise the poems from this point of view the truth can be
recovered from their own words.
Most readers of the Iliad will agree that the scenes in Troy
have an un-Greek air about them. In contrast to Agamemnon
and his peers, who are marshals of men, good at the war-cry,
matchless in sword-thrust and spear-throw, Priam is a mild
and gentle ruler, anything but warlike. Hector, it is true,
might pass for an Achaean, but Paris, who seems to be an older
figure in the saga, 18 is notoriously unmanly. Similarly, while
the status of Andromache as Hector’s wife is not very different
from Penelope’s, Helen is free to go about the streets, thrilling
the onlookers with her beauty, and Hecuba is a personage of
great dignity and influence. It is she, not the King, who directs
the act of intercession for the safety of the city:
Hecuba went into the palace and ordered her ladies-in-waiting to gather
the old women together from all parts of the city. Then she entered her
bedchamber, where she stored her embroidered robes, woven by the women
of Sidon that Paris had brought back with him on the voyage on which he
had fetched Helen; and one of these, the largest and most richly adorned,
lying at die bottom of the pile like a duster of brilliant stars, she picked out
14 11 3. 72, 7. 362-4. is Scott UH 205-6.
XIII
THE CLASH OF CULTURES
4*7
as a gift for Athena. The old women thronged after her to the shrine in the
citadel, and the door was opened to them by Theano, daughter of Kisseus
and wife of Antenor, whom the Trojans had chosen to be priestess. They all
lifted up their hands with cries of Alleluia! and Theano took the robe and
laid it on the goddess's lap. 18
The same goddess was of course served by priestesses at Athens
and elsewhere throughout her history; but in later times all her
public services — sacrifices, processions, games — were controlled
by the state officials, who were men. This Trojan Athena is
what the Athenian Athena ceased to be.
Priam’s palace was arranged as follows:
In it were fifty bedchambers of polished stone, where his sons slept with
their wives beside them; and on the other side of the court were twelve
more, built close together of polished stone, where his daughters slept with
his sons-in-law. 17
This menage has been characterised as a patriarchal joint
household, 18 but a little reflection shows that such an inter-
pretation is impossible. The patriarchal joint household con-
sists of the paterfamilias, his wife, sons, unmarried daughters,
and daughters-in-law. It does not include the married daughters
or the sons-in-law, because they live in other households. An
establishment like Priam's must from the nature of the case
have been exceptional, because the married children cannot
have lived in two houses at once. It is exceptional because it is
royal. It is constituted on the principle of matriarchal en-
dogamy, which enables the sons to secure the succession
by marrying their sisters. That this had once been the rule
in Priam’s city is proved by die arrangement of his palace.
The Homeric poets have described it without tinderstanding
it. But they were acquainted with the practice. Aiolos, the
king of the winds, lived on a magic island in a palace en-
compassed with a wall of bronze:
Here twelve children were bom to him, six sons and six daughters, and he
gave his daughters in marriage to his sons. All day long they feasted with
their father and mother on the abundance of good things that were set
before them, while the smoke curled up and the courtyard rang, and at night
they slept under counterpanes with their wives beside them. 18
18 II. 6. 286-303.
17 ll. 6. 243-50. Nestor’s palace at Pylos was of the same type: Oi. 3.
387, 451. 18 Erdmann 126. 18 Oi. 10. 3-12.
Cc
418 STUDIES IN ANCIENT GREEK SOCIETY XIII
In the world of myth the incestuous customs of the banished
past survive, because, being divorced from reality, they are in-
nocuous.
We have already been introduced to the marvellous kingdom
of Phaeacia. Although mythical, it is compounded of materials
drawn from the real world, and we have only to look
into it to see what that world is. It is the vanished
world of Minoan Greece. The royal gardens,
watered by fountains, in which the grapes
are being pressed in one corner while a
fresh crop blossoms in another; the palace
itself with its bronze gates and friezes of
cyanus, guarded at the entrance by hounds
of gold and silver; the flashing feet of the boy
dancers, ^the purple ball thrown into the air
and caught at a leap in time with the music,
while the minstrel sings of the amours of
Ares and Aphrodite, languorous and seductive;
and the free, independent bearing of the
women — these might be scenes from frescoes
on the walls of the Palace of Minos,
gives Odysseus careful instructions how to
FIG. 63. Mycenean
lady: fresco from
Tirytts
Nausikaa
approach her parents:
When you enter the palace, walk straight across to my mother. You will
find her by the fire spinning sea-purple wool, with her chair against the
pillar and her serving-women at her side. My father will be sitting there too,
sipping his wine like an immortal, but pass him by and clasp my mother’s
knees — then, however far away, you may be sure of a safe journey home. 20
The king is only a decorative figure. The queen will decide.
This queen is a remarkable woman. In the streets Odysseus
meets a girl with a pitcher, who shows him the way to the
palace and gives some information about the royal family:
Alkinoos took her to wife, and He honours her as no other woman is
honoured by the husband for whom she keeps house in all the world to-day;
and not only he but her children and the whole people — they look on her
as a goddess when they salute her as she passes through the streets. So
shrewd and sensible is she that she even composes disputes among the men.
If you win her heart, you will have good hope of returning to your own
country and setting eyes once more on your kith and kin, 2 *
20 Od. 6. 303-15. 21 Od. 7. 66-77.
XIII THE CLASH OF CULTURES -419
So far all is of a piece, but when Odysseus reaches his destina-
tion things turn out quite differently:
Odysseus walked through the house, invisible in the mist in which
Athena wrapped him, till he came to King Alkinoos and Queen Arete.
Just as he clasped the Queen's knees, the mist lifted, and everyone looked
at him in amazement. 'Arete', he cried, 'daughter of Rhexenor,- here at
your feet I beseech you, after all my sufferings, you and your guests, on
whom may Zeus bestow health and wealth while they live and children to
inherit their possessions and the prerogatives they have received from the
people — after all I have endured on my wanderings, grant me safe conduct
back to my native land!' With these words he sat down in the cinders on the
hearth, and still the company was silent. At last Echeneus, the oldest of the
Phxacians, a man with an eloquent tongue and a head full of ancient
wisdom, spoke to the King and said: ‘Alkinoos, why do you keep the
stranger sitting in the cinders? We are all waiting for a word from you. Raise
him up and lead him to a chair; order the serving-men to mix the wine so
that we may offer a grace to Zeus who walks in the suppliant’s footsteps, and
tell the housekeeper to give him supper.' Alkinoos then took the stranger's
hand and led him to the chair next his own, which Laodamas, his favourite
son, had vacated for him. 22
Odysseus has addressed his petition to the Queen, as he was told
to do, but she does not answer. It is the Kingwho decides after all.
What we have here are two renderings of the same theme. In
one, the suppliant places himself at the queen’s feet and clasps
her knees — a symbolic gesture of birch, rebirth, adoption, sup-
plication. In the other, he sits down at the hearth — the centre
of family life — and is led by the king to a place at table and
thereby accepted as a kinsman. Both are rites of supplication,
but the one is matriarchal, the other patriarchal, and the poets
who put the Odyssey in its final shape hesitated between the two.
When we enquire into the antecedents of this royal pair, a
further anomaly is brought to light. Again it is the girl with the
pitcher speaking:
The first person you will meet when you enter the palace will be the mis-
tress of the house, Arete, bom of the same parents as Alkinoos. Poseidon
had a son, Nausithoos, by Periboia, the youngest daughter of Eurymedon,
King of the Giants, The Giants were a foolish people and perished king and
all. Nausithoos reigned over the Phxacians and begot two sons, Rhexenor
and Alkinoos. Rhexenor was slain soon after marriage by Apollo, leaving
an only child, Arete, whom Alkinoos took to wife. 2 *
, 22 Od. 7. 1 39-71. 22 Od. 7. 53 66: see Roscher LGRM 3. 2206-7,
J. A. K. Thomson 168, Kagarov PPK 37.
420 STUDIES IN ANCIENT GREEK SOCIETY XIII
At the end of this passage we are told that Arete is a daughter
of her husband’s brother. This is patriarchal endogamy, cor-
responding to the Attic law of the heiress (p. 139). Arete was an
only child, so she was married to her father’s next-of-kin.
And this is only a few lines after we have read that Alkinoos
and Arete were ‘born of the same parents’. Scholars have tried
to save the situation by rendering the word for ‘parents’ as
‘ancestors’, but this only brings them into conflict with Hesiod,
who described the couple as brother and sister. 24
Here, if anywhere, Homer nods. He has been caught in
flagrante delicto. A pedigree conforming to the familiar and
respectable law of the heiress has been invented as an alternative
to the rule of matriarchal endogamy, which the Greek poets
did not understand and did not wish to understand. A mere
slip of the tongue has betrayed them. There can be no doubt
about the verdict, and its implications are disquieting. How
many other pedigrees have they tampered with for similar
motives but with greater skill? Even in this case, palpable
though it is, they have got away with it for nearly three mil-
lennia. That is because, in matters affecting the social status of
the sexes, their modern editors, for all their scholarly detach-
ment, also nod.
3. The Kingdom of Odysseus
Having lifted a corner of the veil, we can discern in outline
the whole picture shining through. If we follow Odysseus back
to Ithaca, we shall see the confusion that resulted when the
hard-headed Achaean corsairs broke into this exotic matriarchal
world.
The obscurities and inconsistencies in the Homeric account
of life in Ithaca are due partly to the poets, who were very im-
perfectly acquainted with the state of society enshrined in the
tradition, but some of them must be put down to the conditions
themselves, which were hill of transitional anomalies arising
from the collision and combination of two different cultures.
The kingdom of Odysseus, inhabited by the Kephallenes, is
24 Hes. fr. 95 = 0 J. 7. 54 sch. The discrepancy was pointed out by
Burrows 217.
XIII THE CLASH OF CULTURES 421
defined as the three islands of Ithaca, Samos (the later Kephal-
lenia) and Zakynthos, together with Krokyleia and Aigilips,
which were either townlands on the coast of Akarnania or
islets in the bay. 26 Ithaca is described as rugged, thickly wooded,
with only rough pasturage, unsuitable for horse-breeding. 20
The wealth of Odysseus consists mainly of livestock, most of
which is on the mainland. The inventory is as follows. On the
THE KINGDOM OF ODYSSEUS Map IX
mainland he has 12 herds of oxen, 12 flocks of sheep, 12
droves of swine, and 12 herds of goats. On the island, despite
the gluttony of the suitors, he has 1 1 herds of goats, 600 sows
and 360 boars, and a temenos of cornland. 27 All these are tended
for him by an unspecified number of slaves, one of whom,
Eumaios , figures prominently in the story. He is a son of
26 II. 2. 631-5, Str. 45Z-3. 20 Od. 4. 601-8, 13. 242-7.
27 Od. 14. 13-20, 100-4, 17. 299.
422 STUDIES IN ANCIENT GREEK SOCIETY XIII
a king of Syros, one of the Cyclades. Kidnapped in early child-
hood by Phoenician pirates, he was purchased from them by
Odysseus' father, Laertes, and brought up with the family . 28
Before the war Odysseus had intended to give him a house, a
holding, and a wife— rin other words, to settle him as an inde-
pendent cultivator . 29 There are also a number of female slaves —
twelve employed in grinding corn and over fifty in the palace . 30
One of the latter, Aktoris, was part of Penelope’s dowry . 81
The housekeeper, Eurykleia, who had nursed Odysseus, was
bought by Laertes for twenty oxen. Laertes was in love with
her but abstained from sexual relations out of respect for his
wife . 83
Odysseus’ mother, who came from Phokis, is dead, but
Laertes is still alive. In Book XXIV, which is generally re-
garded as one of the latest, the old man recalls how he had once
led a raid against the mainland and speaks of himself as having
been king at the time . 38 But there is no mention of this in the
rest of the poem, nor are we told why, if he had been king, he
had abdicated. The uncertainty on this point suggests that
Odysseus may have become king by some other means than
succession to his father.
Laertes does not live in the palace, nor is it said that he had
ever done so. He lives at some distance on a country estate,
including a house, a threshing-floor, and a garden . 34 In Book
XXIV it is described as his own property acquired with much
labour — that is, an assart or clearing which he had made for
himself in the waste . 38 The old man spends his time pottering
round the garden, the land being worked for him by Dolios
and his family . 38 This Dolios is a slave given to Penelope by
her father, Ikarios, when she married and came to Ithaca . 87 Like
Aktoris, he was part of her dowry. In Book IV she tells her
serving-women to go and fetch ‘old Dolios, my slave, who was
given to me when I came here by my father and keeps my
garden’. To whom does this garden belong? Again we see that-
28 Od. 15. 403-84, 363-70.
20 Od. 14. 62-4, cf. 17. 320-3; Vinogradoff GM 202, 230, Bloch 239.
32 Od. 20. 105-8, 22. 421-3. 81 ol 23. 228. 82 od. 1. 429-35.
33 od. 24. 377—8. 34 Od. 1. 189-93, 23. 359-60. 86 Od. 24, 205-7.
30 od. 24. 220-5, 387-90, 497. 37 Od. 4. 735-7.
XIII THE CLASH OF CULTURES 423
Book XXTV is out of step with the rest of the poem. From
Book IV it appears that the dowry had included the land itself
as well as the slave that worked it.
Odysseus sailed to Troy with a contingent of twelve ships.
From incidental allusions it can be calculated that their com-
plement amounted to 624 men — that is to say, each vessel
was manned by a captain, coxswain, and fifty oarsmen. 38 The
population has thus been seriously depleted. Not only is the
king himself abroad, but he has taken with him what must
have been a substantial portion of the adult males. And that was
twenty years ago. During all these years there has been no
acknowledged ruler, no council of elders, no meeting of the
assembly. 88 And meanwhile a new generation has sprung up,
including a large number of ambitious young men who like
Telemachos have never known their fathers. They are free from
(he restraints which in normal times would have been imposed
on them by their elders, and not unnaturally, assuming that
Odysseus is dead, they are impatient to cut their losses and
appoint a new king.
It is from these young men that the suitors are drawn. There
are 108 of them — 12 from Ithaca, 24 from Samos, 20 from
Zakynthos, and 52 from Doulichion. 40 For three years they
have been pestering Penelope, 41 feeding and idling in the house
at Telemachos’ expense. Their intrusion seems to rest on some
undefined claim to the hospitality of the royal table, such as
would have been accorded in normal times to the elders. 42
They enjoy at least the passive support of the people, 43 and
refuse to leave till Penelope marries one of them. Telemachos
cannot get rid of them; he can only insist on his right to inherit
38 II. 2. 637, Od. 9. 6o-i, 159-60, 195, 289, 311, 10. 1x6, 128-34,
203-8.
38 Od. 2. 26-7.
48 Od. 16. 247-51. Where Doulichion was is not clear. Ancient authorities
identified it with Kephallenia (Str. 456); in recent times it has been
equated with Leukas (Allen HCS 83-7) and with the islands between
Ithaca and the mainland (Rodd 78-97). Its inhabitants are described as
Taphioi: E. IA. 283-7. In the Iliad it belongs to Meges, whose father had
come from Elis (2. 625-9) hut in the Odyssey it is ruled by Akastos, of
unknown parentage, and Meges is not mentioned (14. 335-6).
41 Od. 13. 377-8. 42 Glotz CG 55. 43 Od. 2. 239-41, cf. 16. 375.
424 STUDIES IN ANCIENT GREEK SOCIETY XIII
his father's house and property. 4 * This they acknowledge.
One of their motives in plotting his death, which would mean
the extinction of the family, is the hope of dividing his
patrimony among themselves. 46 On the other hand, Tele-
machos pointedly refrains from claiming the succession to the
kingship. He admits that there are other princes in the islands
who may succeed his father. 46 What then do the suitors hope to
gain by marrying Penelope? The answer to this question, on
which the whole story turns, is nowhere definitely stated. It is
just taken for granted. But the situation is such as to allow of
only one answer, which slips out incidentally in Book XV.
Telemachos says that the aim of the suitors is 'to marry my
mother and possess my father’s prerogative (g/nw)’. 47 Whoever
wins Penelope will succeed Odysseus. The kingship is not
hereditary in the male line; it goes with the hand of the queen.
The moment selected by Odysseus for his attack on the
suitors is when they are engaged in the archery contest. Two
rows of axes have been set up in the hall, and Penelope has
promised that whoever succeeds in stringing her husband’s
bow and shooting an arrow between the lines of axes shall have
her to wife. This is another example of the pre-nuptial con-
test (p. 404), and it may be noted that in a post-Homeric
tradition Odysseus himself won the hand of Penelope by
beating his rivals in a foot-race arranged by her father. 48 In
Book II the suitors urge Telemachos to send Penelope home to
her father, who will arrange a second match for her. 49 In
Book XV Ikarios himself and her brothers are pressing her to
marry Eurymachos, one of the leading suitors. 60 We are not
told where her father lives, but it seems from this passage that
he cannot be far away. In later literature he was domiciled at
Sparta, but that cannot have been intended in the Odyssey, or we
should have heard about him when Telemachos went there.
The silence of the Odyssey on this point can be supplemented
from a tradition cited by Strabo from the later epics. Ikarios
and Tyndareos were brothers, born and bred at Sparta. In
early manhood they had to fly the country, and they took
refuge with Thestios, the king of Pleuron, which lies at the
44 Od. 1. 397-8. 46 Od. 2. 335-6, 368. 4 <> Od. 1. 394-6.
47 Od. 15. 518-22. 48 Paus. 3. 12. i. 49 0<i.2. 113-4. 60 Od. 15. 518-22.
XIII THE CLASH OF CULTURES 425
entrance to the Gulf of Corinth. Tyndareos eventually returned
to Sparta, where he became king, but Ikarios remained in the
north and received the kingdom of Akamania, which he ruled
jointly with his sons, Alyzeus and , Leukadios. 61 These, then,
are the brothers of Penelope mentioned in the Odyssey. Alyzeus
stands for Alyzia, a town in Akamania, and Leukadios for
Leukas. Since these are close to the territory of Odysseus as
H f fin p d in the Odyssey, we are led to conclude that he ruled as a
vassal of Ikarios by right of marriage to his’ daughter.
There is another version of the marriage of Odysseus — a
local tradition from Sparta. After the wedding Ikarios begged
him to make his home with him, but Odysseus declined, and,
when the couple left for Ithaca, the importunate father-in-law
followed them. Eventually Odysseus turned to his bride and
told her to take her choice — to go with him to Lis own home
or return to her father’s without him. 62 This is a clear folk-
memory of the transition from matrilocal to patrilocal mar-
riage. It will be recalled how, after living with his wives’
people for twenty years, fourteen of which he spent in service
to mem, Jacob ran away with them to his father's house, pur-
sued by the indignant Laban, who realised that he was losing
control of his daughters' property. 63 So with Penelope: she
chose to leave her parents and cleave unto her husband.
We see then that the Odyssey is full of vague memories of the
tensions and contradictions that marked the transition from
mother-right to father-right; and this raises the question, how
far can the persons and peoples of the story be identified with
the ethnical groups known to have been active in this formative
period of Greek history?
4. The Leleges of Western Greece
In upbringing, behaviour, and outlook the Homeric Odys-
seus is indistinguishable from the Homeric Achilles. Whether
he was an Achaean in the strict sense cannot be positively
decided. His grandfather, Arkeisios, is mentioned in the
poem, but with no clue to his origin. 64 His family seems to be
61 Str. 452, 461. 82 paus. 3. 20. 10.
63 Gen. 31. x-43, cf. E. fr. 318. 64 Ol 13. 182, 16. 1x8.
426 STUDIES IN ANCIENT GREEK SOCIETY XIII
somewhat isolated in the islands, with no blood relatives out-
side its own circle. Even Mentor, whom he left in charge of his
domestic affairs when he went to the war, is only an old friend
of the family . 58 The only person, apart from his immediate
relatives, to whom he refers as a kinsman, is Eurylochos, the
second-in-command of his contingent . 56 According to the
ancient commentators Eurylochos was the husband of his
sister Ktimene, whose marriage is mentioned in the poem as
having brought the family a substantial bride-price . 57
There are some traditions relating to the Kephallenes which,
though not mentioned in the Odyssey, agree with what we are
told there and have an important bearing on Ikarios. The
people of Ithaca used to draw water from a well just outside
the town. It lay in a sacred grove of poplars near a wayside
shrine of the nymphs. It was built by Polyktor, Neritos, and
Ithakos. This information is given in the Odyssey. 68 The trio
were evidently local heroes. Polyktor is otherwise unknown to
us, except that a man of the same name was the father of one
of the suitors, implying that his descendants were still in
the islands . 50 Neritos is the eponym of the mountain in die
middle of Ithaca, Ithakos of the island itself. It seems that the
poets of the Odyssey knew more about these ancient figures
than they have recorded in the poem. From other sources we
learn that Neritos and Ithakos were brothers, born in Kephal-
lenia. Their father was Pterelaos, who had two other sons,
Taphios and Teleboas . 60 The Taphioi and Teleboai were
brigands in occupation of the small islands scattered between
Leukas and the Gulf of Corinth. The former are mentioned
several times in the Odyssey; the latter, according to Apol-
lodoros, were visited on one occasion by a punitive expedition
from Mycenae . 61 Both came from Akarnania, and were treated
by some authorities as one and the same people . 63 The Taphioi
55 Od. 2. 225-7. 86 Od. 10. 205, 44.1. 87 Od. 10. 441 sch., 15. 363-7.
56 Od. 1 7. 204-11. 50 Od. 18. 299.
60 Acus. 30, A.R. 1. 747 sch. The pedigree is confused: see Roscher
LGRM 3. 3261-2.
01 Od. 1. 105, 181, 419, 14. 452, 15. 427, 16. 426, Apld. 2. 4. 6.
02 Str. 461, A.R. 1. 747 sch., Hsch. Tntep6cn. Telebois was the old name
of Akarnania (St. B. Tn^pots) and Teleboai of the Taphioi (Str. 459).
XIII
THE CLASH OF CULTURES
427
were also described as Phoenicians or Leleges. 6 ® A Phoenician
origin can be ruled out. There is no sign of it in the Odyssey,
and it may have been suggested by a passage in Book XV,
where they are said to have raided the coast of Phoenicia.® 4
The alternative has die support of a tradition preserved by
Aristode. In this Taphios and Teleboas are sons of Hippothoe
by Poseidon. Hippothoe is given by Apollodoros as a grand-
daughter of Perseus, but according to Aristotle her father was a
‘son of the soil' from Leukas and his name was Lelex. 66 The
ttvo versions are not incompatible. Perseus came from the
Cyclades; the pirates expelled from the Cyclades by Minos
were Carians and Leleges. The inference is that the Taphioi
and Teleboai, and with them the Kephallenes, were Leleges
who after being driven from the ASgean to the Adriatic
hovered like hornets round the approaches to the Gulf of
Corinth.
This conclusion is confirmed by topography. Apart from
Ithake (Ithaca) and Astakos on the coast of Akarnania, place-
names in -akt and -akos do not occur in Greece at all; in Anatolia
they are common. 68 The river Euenos, which reaches the sea
near Pleuron, has a namesake near Troy. 67 Samos, the Homeric
name for Kephallenia, corresponds to the Aegean Samos,
which was one of the oldest settlements of the Leleges (p. 1 66 );
and one of the sons of Ankaios, who ruled the Leleges of the
Aegean Samos, was Alitherses, which is the name of the old
prophet of Ithaca in the Odyssey .° 8
I have told the story of Ikarios. Born at Sparta, he fled with
his brother to Pleuron, where he was received by Thestios and
became king of Akarnania. That was one version, but there
®3 EM. 748. 40. 64 Oi. 15. 425-7. 05 Apld. 2. 4. 5, Arist. fir. 546.
66 1 exclude Phylake, which is a Greek word. Examples: Artake, Rhyndakos,
Chabake, on the Black Sea; Idakos, Andriake, in Thrace; Acharake in Lydia;
Mazaka in Cappadocia; Symbake in Armenia. The personal names Assarakos
and Hyrtakos are also Anatolian.
67 Str. 327, 614.
6 ® Pans. 7. 4. i, Oi. 2. 157-9. Heurtley found an abundance of Early
Helladic pottery among the ruins of Pelikata, which he identified as
the palace of Odysseus, and inferred that ‘in Ithaca Minyan and Mycenean
influences were only thinly spread over an earlier civilisation, which con-
' tinued to survive’: SPO 414.
428 STUDIES IN ANCIENT GREEK SOCIETY XIII
were two others. The Spartans denied that the brothers had
ever left Sparta. One of their kings, Perieres, married a
daughter of Perseus, and had four sons — Ikarios, Tyndareos,
Hippokoon, and Oibalos. When he died they disputed the
succession. Supported by Ikarios, Hippokoon expelled Tyn-
dareos, who fled to Pellana a few miles higher up the Eurotas.
Table XVI
PERIERES AND THESTIOS
Arcs
Perieres=Gorgophone Thestios
Ikarios Aphareus Tyndareos = Leda = Zeus
I ) ,1 I
Penelope Idas= Clytemnestra Helen
■= Odysseus Marpessa =Agamem- ■= Menelaos
non
Then Hippokoon was slain by Herakles, and Tyndareos was
restored. He became king and married Leda, a daughter of
Thestios . 89 It may be doubted, however, whether, at least in
early times, the Spartans regarded Thestios as king of Pleuron
beyond the Corinthian Gulf, because they remembered him. as
the founder of a village called Thestia on the banks of their own
Eurotas . 70 The Messenian version was different again. They
said that Tyndareos fled to Aphareus, another brother, who
installed him at Thalamai, where he married Leda. We' are not
told where Leda had come from, but Idas, a son of Aphareus,
carried off the lovely Marpessa, whose father was Euenos, a
brother of Thestios and a native of Pleuron . 71
It is meaningless to ask which of these versions is the correct
00 Paus. 3. 1. 4-5.
Cedr. Hist. Comp. 2i2=Migne 121. Mai. Chon. 82=Migne 97. 164.'
71 Apld. 1. 7. 7-8. An indication of matrilineal succession survives
in the saga of Melcagros, who killed his mother’s brothers, the sons of
Thestios, because they took from him the hide of the Calydonian boar,
which he wanted to give to Atalanta (a hypostasis of Artemis), on the ground
that if he was not going to keep it for .himself it belonged to them 'by right
of birth’: Apld. 1. 8. 2-3.
Euenos
Marpessa
=Idas
XIII
THE CLASH OF CULTURES 429
one. They are all fictions. Yet they convey a historical truth,
which manifests itself through their contradictions. If this
myth had been systematised, as so many were, its m eaning
would very likely have been irrecoverable. As it stands, it is
self-revealing. We have it in three versions belonging to three
localities, and so the problem is to identify its provenance in
such a way as to account for its distribution.
Besides their settlements in Leukas and Akarnania the
Leleges had another further up the Gulf in Lokris Ozolis and
yet another at its head. Lelex was one of the early kings of
Megara. His grandson, Pylas, led a band of Leleges from there
into Messenia, where he founded Pylos. Expelled from thence
by the Neleidai, they moved up the coast and founded the
other Pylos in Triphylia . 72 Lelex was also the first king of
Sparta, whose earliest inhabitants were Leleges. One of the
town wards, which shared in the worship of Artemis Orthia,
was Pitana. This is the only settlement in metropolitan Greece
named after an Amazon — the same who founded Pitana
(Pitane) in Aiolis. 7 * In Chapter VII we traced the cult of
Artemis Orthia to Ephesos, whence it was brought to the
Peloponnese by Carians and Leleges.
Leda had two daughters, Clytemnestra and Helen. Cly-
temnestra’s father was Tyndareos, but Helen sprang from an egg
laid or found by Leda after she had been visited by Zeus in the
guise of a swan . 74 This totemic myth is the kernel of the whole
tradition. The name Tyndareos is not Indo-European. The com-
bination -nd- is alien to Greek except as the result of composition
or contraction. It is specially common in Caria — Lindos, Myndos,
Karyanda, Alabanda, etc. And who is Leda — Lada in Doric — but
another form of Leto, lada, the Carian 'woman' (p. Z94)? 76
72 Paus. 4. 36. j. 72 Paus. 3. 16. 9, D.S. 3. 54, Held. Pont. 34.
74 E. Hel. 16-22, Sapph. 105, Apld. 3. 10. 7.
7 ® Krappe 363. Leda's association with the swan reminds us that water-
fowl figured in the cults of Artemis Limnaia at Sparta and Stymphalos:
Harrison T 1 14, Imhoof-Blumer NCP 99. Penelope bears the name of a
water-bird (irTjvIXovy). She is said to have been buried at Mantineia (Paus.
8. 12. 5) and to have been the mother of the Arcadian god Pan: Pi. fr. 422,
Hdt. 2. 145. 4, cf. Paus. 8. 14. 4-5. This Arcadian Penelope derives from
cults of Artemis Limnaia (J. A. K. Thomson 48-50) introduced by Carians
and Leleges (p. 273).
430 STUDIES IN ANCIENT GREEK SOCIETY XIII
Clytemnestra' and Helen, the wives of Agamemnon and
Menelaos, bring us back to the Pelopidai. The domains of this
dynasty were three — the district between Mycenae and the
Isthmus, the Peloponnesian Achaia, and Laconia with southern
Messenia. The first was the old Mycenean kingdom of the
Perseidai. Their occupation of Achaia was perhaps dictated by
the need to keep the Gulf clear of pirates. The third was ruled
from the palace of Menelaos at Therapne near Sparta, and
Therapne was a daughter of Lelex. 70 For these reasons we may
suppose that Menelaos secured Laconia by marrying into the
reigning dynasty of Leleges, just as Odysseus married the
daughter of Ikarios.
5. The Superiority of the Achaans
Behind the work of the humane poets who composed the
Iliad and Odyssey lies an age of brutality and violence, in which
the bold pioneers of private property had ransacked the
opulent, hieratic, sophisticated civilisation of the Minoan
matriarchate. The old world of Phasacia is doomed; the future
lies with rugged Ithaca. The poets transmuted Odysseus, but
for all that he remains, like Achilles, the ideal of heroic man-
hood. Restless, cunning, enterprising, he has spent ten years at
the wars, and ten more travelling, trading, plundering, piling
up riches, refreshing himself in the arms of a Circe or a Cal-
ypso; and meanwhile, besieged by self-seeking suitors, Penelope
has turned a deaf ear on them all, slaving away at her loom and
waiting submissively for her lord’s return;
Go -to your room and mind your own business, your loom and distaff, and
tell your maids to get on with their work. Talking is men’s business, and
mine above all, because I am master of the house. 77
She is not so well placed as Arete was. Penelope is the pattern
of heroic womanhood, consciously contrasted with Clytem-
nestra, who fought back and died rather than give in. After
ten years Agamemnon returns with a concubine, and nobody
has a word to say against him. Clytemnestra has consoled
herself with a paramour, but that is a different matter. The
heroine of the old order is the criminal of the new. ASschylus,
70 Paus. 3. 19. 9. 77 Od. 1. 356-9.
XIII
THE CLASH OF CULTURES
431
the great poet of the new order, justified the subjection of
woman, but there was something in him, inherited perhaps
from his ancient family traditions, that rebelled; and this,
driven down into his subconscious, re-emerged as an imagina-
tive symbol of the conflict in the magnificent figure that
dominates the masterpiece of Greek poetry .’ 8
The basic unit of mature Greek society, in which succession
passed from father to son, in which the wife was bound to one
man while the husband was free, had been imposed after a
protracted struggle on an entirely different system, in which
succession had followed the female line, in which there had
been no formal matrimony, and the woman had mated as she
pleased. And this struggle had been carried through by the
people who forged, as one of their most effective instruments
for consolidating the new patriarchal ideology,, the epic tradi-
tion embodied in the Iliad and Odyssey . It is this historical
factor that gives the poems their dynamic vitality.
The decline and fall of Minoan civilisation has been com-
pared to the decline and fall of Rome. In both an advanced
but effete society collapsed under the impact of barbarian
invasions; in both the conquerors absorbed the culture of the
conquered and evolved a specific form of poetty, the epic; in
both the eventual result was to produce a new society of a
higher type. But of course there were fundamental differences,
and one of these has never been explained. The Germanic
nations that settled in the Roman provinces adopted the Latin
language; the Achaeans not only preserved their own language
but imposed it on the conquered. In this respect their achieve-
ment has more in common with the Aryan conquest of India,
which spread the speech of the invaders step by step with the
break-up of the pre-Aryan matriarchal cultures . 79 One of the
The murder of Agamemnon has the same social significance as the
crimes of the Lemnian women (p. 175) and the Danaids (see my AA
299). Among the Panwar-Rajputs, the bride-to-be receives a formal homily
in verse from her mother, who condudes by instructing her, if she cannot
persuade the bridegroom to reside with her according to the matriarchal
rule, to poison him (Russell 4. 344). This is now treated as a joke, but at
one time, as Ehrenfels remarks (142), it was 'by no means mere irony but
practical life.'
v® Ehrenfels 138.
432 STUDIES IN ANCIENT GREEK SOCIETY XIII
major factors enabling the Achaeans to impose Greek speech was
their form of social organisation, which, being adapted to the
growth of private property, was charged with immense
historical potentialities, and at the same time, being patri-
archal, it turned the scales, wherever Greek was brought by
intermarriage into collision with the old vernaculars, in favour
of the men and the language of the men.
To what extent they had succeeded in this task before the
coming of the Dorians, who completed it, we do not know,
because it is impossible to say how far the Homeric portrait
of them has been coloured retrospectively; but the fate of
Orestes suggests that in the days of the Pelopidai the paternal
title was still insecure. And what is more, when his sons and
grandsons fled across the AEgean, they found themselves back
again in the old matriarchal world. The struggle had to be
fought out afresh. Transplanted to Aiolis and Ionia, the
Achaean art of minstrelsy entered on a new phase of growth
and slowly ripened into epic.
Greek civilisation did not descend on peaceful valleys like
Iris from Olympus. It was the fruit of struggle, fought for in
innumerable raids and battles amidst the smoke of burning
cities and the groans of homeless captives. The force that drove
it forward was the class-struggle. To overlook this is a poor
compliment to the Greeks and a disservice to ourselves.
Part Five
HOMER
The relation between formal music and speech will yet
become the subject of science, not less than the occa-
sion of artistic discovery.
YEATS
XIV
THE ART OF POETRY
I. Speech and Magic
The subject of this chapter is the origin and nature of poetry,
and it will be treated as a scientific problem. To those who are
content to enjoy poetry for its own sake this approach
may seem inappropriate or unattractive; but studied scienti-
fically poetry is more, not less, enjoyable. To enjoy it fully
we must understand what it is, ana to understand what it is
we must enquire how it has come into being and grown up.
Our object in raising this problem is to seek light on the
prehistory of Greek poetry, but it can only be solved by col-
lating material from as wide a field as we can. Accordingly,
our examples will not be confined to Greek poetry. I shall draw
freely on English poetry, which is useful because it is the
most familiar, and on Irish poetry, which illustrates an earlier
stage in the development of modem European poetry, and also
on the songs and dances of primitive peoples.
One of the most striking differences between Greek and
modern English poetry is that in ancient Greece poetry was
wedded to music. There was no purely instrumental music,
and much of the finest poetry was composed for musical
accompaniment. In Irish too there is a close union between
poetry and music, and here it is not just a matter of inference.
It is still a living reality. I shall never forget the first time I
heard some of the Irish poems I had long known in print sung
by an accomplished peasant singer in the traditional style. It
was an entirely new experience to me. I had never heard any-
thing like it, in poetry or music.
Irish poetry has another characteristic. To most English
people English poetry is a closed book. They neither know nor
care about it. And even the few that take an interest in it —
there are not many even of these of whom it can be said that
poetry enters largely or deeply into their daily lives. Among
436 STUDIES IN ANCIENT GREEK SOCIETY XIV
the Irish-speaking peasantry it is different. For them poetry
has nothing to do with books at all. Most of them are, or were
till recently, illiterate. It lives on their lips. Everybody knows
it. Everybody loves it. It is constantly bubbling up in everyday
conversation. And it is still creative. Whenever a notable
event occurs, a song is composed to celebrate it. I say com-
posed, but the word, is hardly applicable. These songs are not
composed in our sense of the word. They are improvised. In
many Irish villages there was till recently a trained traditional'
poet, who had the gift of producing poems, often in elaborate
verse forms — far more elaborate than ours in modern English —
on the inspiration of the moment. In the village I knew best
there was a famous poet, who died about forty years ago. His
poems were nearly all improvised and occasional. I remember
being told by his family how on the night he died he lay in
bed with his head propped on his elbow pouring out a con-
tinuous stream of poetry.
This man was of course exceptionally gifted. He was a pro-
fessional poet, who had learnt his craft under some poet of the
preceding generation. But I soon found that no sharp line
could be drawn between the professional poet and the rest of
the community. It was only a matter of degree. To some extent
they were all poets. Their conversation is always tending to
burst into poetry. Just as extant poetry is more widely known
than it is in our society, so the ordinary person is something of
a poet. Let me give an example.
One evening, strolling through this village perched high up
over the Atlantic, I came to the village well. There I met' a
friend of mine, an old peasant woman. She had just filled her
buckets and stood looking out over the sea. Her husband was
dead, and her seven sons had all been 'gathered away’, as she
expressed it, to Springfield, Massachusetts. A few days before
a letter had arrived from one of them, urging her to follow
them, so that she could end her days in comfort, and promising
to send the passage money if only she would agree. All this she
told me in detail, and described her life — the trudge to the
turf stack in the hills, the loss of her hens, the dark, smoky
cabin; then she spoke of America as she imagined it to be —
an Eldorado where you could pick up the gold on the
XIV
THE ART OF POETRY
437
pavements— and therailway journey to Cork, and the transatlantic
crossing, and her longing that her bones might rest in Irish
soil. As she spoke, she grew excited, her language became more
fluent, more highly-coloured, rhythmical, melodious, and her
body swayed in a dreamy, cradle-like accompaniment. Then
she picked up her buckets with a laugh, wished me good night,
and went home.
This unpremeditated outburst from an illiterate old woman
with no artistic pretensions had all the characteristics of poetry.
It was inspired. What do we mean when we speak of a poet as
inspired?
' To answer this question -we must turn to primitive poetry
as it still lives on the lips of savages at the present day. But
we cannot understand the poetry of these peoples unless we
know something about their society. Further, poetry is a
special form of speech. If we are to study the origin of poetry,
we must study the origin of speech. And this means the origin
of man himself, because speech is one of his distinctive charac-
teristics. We must go right back to the beginning.
We are still a long way from understanding folly how man
came into existence, but there is one fundamental point on
which scientists are agreed. Man is distinguished from the
animals by two main characters — tools and speech.
The primates differ from the lower vertebrates in being able
to stand upright and use their forefeet as hands. This develop-
ment, involving a progressive refinement of the motor organs
of the brain, arose from the special conditions of their en-
vironment. They were forest animals, and life in trees demands
dose co-ordination of sight and touch and delicate muscular
control. And once developed the hands presented the brain
with new problems, new possibilities. Thus, from the beginning
there was an integral connection between hand and brain . 1
Man differs from the anthropoid apes, the next highest of the
primates, in being able to walk as well as stand. It has been
suggested that he learnt to walk as a result of deforestation,
which forced him to the ground. Be that as it may, in him the
division of function between hands and feet was completed.
His toes lost their prehensility; his fingers attained a degree of
1 Elliot Smith 17-46, Clark 1-6.
438 STUDIES IN ANCIENT GREEK SOCIETY XIV
dexterity unknown among the apes. Apes can manipulate
sticks and stones, but only human hands can fashion them into
tools.
This step was decisive. It opened up a new mode of life.
Equipped with tools, man produced his food instead of merely
appropriating it. He used his tools to control nature. And in
struggling to control it he became conscious of it as something
governed by its own laws, independent of his will. He learnt
how things happen, and so how to make them happen. As he
came to recognise the objective necessity of natural laws, he
acquired the power of operating them for his own ends. He
ceased to be their slave and became their master. 2 On the
other hand, in so far as he failed to recognise the objective
necessity of natural laws, he treated the world around him as
though it could be changed by a mere assertion of his will.
This, as explained in Chapter I (p. 38), is the basis of
magic.
In its initial stages the labour of production was collective.
Many hands worked together. In these conditions the use of
tools promoted a new mode of communication. The cries of
animals are severely limited in scope. In man they became
articulate. They were elaborated and systematised as a means
of co-ordinating the actions of die group. And so in inventing
tools man invented speech. 3 Again we see the connection
between hand and brain. Speech emerged as part of the actual
technique of production. It assisted the muscular movements
of the/ body by prefiguring the labour process; and being in-
dispensable to that process it appeared subjectively as its
cause — in other words, it was magical. In primitive thought
the spoken word is universally invested with a magical power. 4
As technique improved, the vocal accompaniment ceased
to be a physical necessity. The workers became capable of
working individually. But the collective apparatus did not
. disappear. It survived in the form of a rehearsal, which they
performed before beginning the real task — a dance in which they
reproduced the collective movements previously inseparable
2 Engels DN 279-96, Lenin xi. 247-51.
3 Malinowski PMLP 310, CGM 2. 235.
4 Briffault 1. 14-23, cf. Malinowski CGM 2. 232.
XIV
THE ART OF POETRY
439
from the task itself. This is the mimetic dance as still practised
by savages to-day.
Meanwhile speech developed. Starting as a directive ac-
companiment to die use of tools, it became language as we
understand it — a fully articulate, fully conscious mode of com-
munication between individuals. In the mimetic dance, how-
ever, where it survived as the spoken part, it retained its
magical function. And so we find in all languages two modes of
speech — common speech, the normal, everyday means of com-
munication between individuals, and poetical speech, a medium
more intense, appropriate to collective acts of ritual, fantastic,
rhythmical, magical.
If this account is correct, it means that the language of
poeoy is essentially more primitive than common speech,
because it preserves in a higher degree the qualities of rhythm,
melody and fantasy inherent in speech as such. Of course
it is only a hypothesis, but it is supported by what we
know of primitive languages. In them the differentiation
between poetical speech and common speech is relatively
incomplete.
The common speech of savages has a strongly marked
rhythm and a lilting melodic accent. In some languages the
accent is so musical, and so vital to the meaning, that when a
song is composed the tune is largely dictated by the natural
melody of the spoken words. 5 And further, die speaker is
always liable to break into quasi-poetical flights of fantasy,
like that Irish peasant woman. The first two of these
characteristics cannot be illustrated here, but the last one
can.
A Swiss missionary was once camping in Zululand close to
the Umbosi railway. For the natives the Umbosi railway
signifies the journey to Durban, Ladysmith, Johannesburg—
the journey made year after year by the boys of the kraal, driven
from home by the poll-tax to wear out their youth in the
mines, and by the girls too, who suffer many of them an even
worse fate in the back-street brothels. One of the servants was
tn the camp cleaning the pots, when a train approached, and he
was overheard muttering these words:
5 Schapera BT 282-3, 286, 401-2, Rattray A 245-7, cf. PI. R, jgSd.
44 ° STUDIES IN ANCIENT GREEK SOCIETY XIV
The one who roars in the distance,
The one who crushes the young men and smashes them,
The one who debauches our wives. -
They desert us, they go to the town to live bad lives.
The ravished And we are left alone.®
Here is another artless soliloquy. It is only an old black servant
mumbling to himself, and yet it is poetry. The train catches
his attention. He forgets the pots. Then he forgets the train.
It ceases to be a train and becomes a symbol for the force that is
destroying all he holds most dear. The dumb resentment of his
subconscious being finds a voice. Then the roar of the train
dies away, and he returns to his pots.
Thus, the common speech of these savages is rhythmical,
melodic, fantastic to a degree which we associate only with
poetry. And if their common speech is poetical, their poetry
is magical. The only poetry they know is song, and their
singing is nearly always accompanied by some bodily action,
designed to effect some change in the external world — to
impose illusion on reality.
The Maoris have a potato dance. The young crop is liable to
be blasted by east winds, so the girls go into the fields and
dance, simulating with their bodies the rush of wind and rain
and the sprouting and blossoming of the crop; and as they
dance they sing, calling on the crop to follow their example . 7
They enact in fantasy the fulfilment of the desired reality.
That is magic, an illusory technique supplementary to the
real technique. But though illusory it is not futile. The dance
cannot have any direct effect on the potatoes, but it can and
does have an appreciable effect on the girls themselves. Inspired
by the- dance in the belief that it will save the crop, they pro-
ceed to the task of tending it with greater confidence and so
with greater energy than before. And so it does have an effect
on the crop after all. It changes their subjective attitude to
reality, and so indirectly it changes reality.
The Maoris are Polynesians. So are the islanders of the New
Hebrides. These have a traditional song-form consisting of two
® Junod LSAT 2. 196-7.
7 Bucher 409-10. The potato in question is the Spanish potato ( Batatas
eiuUs).
XIV
THE ART OF POETRY
441
alternating stanzas in different rhythms. The first is termed
the ‘leaf’, the second the ‘fruit '. 8 In Tikopia, another Poly-
nesian island, there is a song-form of three stanzas. The term
for the first means properly the 'base of the tree'; for the second,
the ‘intermediate words'; for the third, the ‘bunch of frait ’. 8
HG. 64. Dancers: Attic vase
The terminology shows that these song-forms have evolved
out of mimetic dances like the dance of the Maori girls.
Poetry has grown out of magic.
Let us carry the argument further. This is one of the in-
cantations collected by Malinowski in the Trobriand Islands:
It passes, it passes,
The breaking pain in the thighbone passes,
The ulceration of the skin passes,
The big black evil of the abdomen passes,
It passes, it passes. 10 ,
The subject of this poem is not what we should call poetical,
but the form is. As Malinowski remarks, the language of these
8 Layard 315. 8 Firth 285. 10 Malinowski CGM 2. 236-7.
442 STUDIES IN ANCIENT GREEK SOCIETY XIV
incantations is distinguished ‘by its richness of phonetic,
rhythmical, metaphorical and alliterative effects, by its weird
cadences and repetitions*. 11 By asserting the truth of what you
wish to be true, you make it come true; and the assertion is
couched in language that echoes the ecstatic music of the
mimetic dance, in which you enacted in fantasy the fulfilment
of the desired reality.
Here is a song from the New Hebrides, addressed to two
women who were said to live in a stone:
The song sings, the song cries,
The song cries, Let her be my wife!
The woman who is there,
The two women, they two
Who are in the sacred stone,
Who sit inside, who live in the stone,
The song cries, Let both come out! 13
Here instead of a statement confusing fact with fancy we have a
command. But the command is not addressed directly to the
persons concerned. It is conveyed through the compelling
magic of the song. The song is externalised as a supernatural
force.
The next example is a German foresters' song:
Klinge du, klinge du, Waldung,
Schalle du, schalie du, Haide,
Halle wider, halle wider, Hainlein,
Tone wider, grosse Laubwald,
Wider meine gute Stimme,
Wider meine goldne Kehle,
Wider mein Lied, das lieblichste!
Wo die Stimme zu verstehen ist,
Werden bald die Biische brechen,
Schichten sich von selbst die Stamme,
Stapeln sich von selbst die Scheiter,
Fiigen sich zum Hof die Klafter,
Haufen sich im Hof die Schober
Ohne jungcr Manner Zutun,
Ohne die gescharften Aexte. 13
The foresters call on the trees to fall to the ground, break up
into logs, roll out of the forest and stack themselves in the
11 lb. 2. 213, cf. 222, Codrington 334, Layard 285, Driberg 245.
13 Layard 142. 13 Bucher 473.
XIV
THE ART OF POETRY
443
yard in answer to their singing. They know very well that all
this is not going to happen, but they like to fancy that it will,
because it helps them in their work. Poetry has grown out of
magic.
My next is from an Old Irish mantic poem:
Good tidings: sea fruitful, wave-washed strand, smiling woods; witch-
craft flees, orchards blossom, cornfields ripen, bees swarm, a cheerful world,
peace and plenty, happy summer. 14
- It was chanted by a prophet as an augury of a good season. The
desired future is described as though already present.
And so by almost imperceptible degrees we reach a type of
poetry with which we are all familiar:
Sumer is icumen in,
Lhude sing cuccul
Groweth sed and bloweth med
And springth the wude nu —
Sing cuccul
The statement here is a statement of fact, but even here it is
accompanied by a command. These seasonal songs, which have
deep roots in the life of the European peasantry, were com-
posed to celebrate the realisation of communal desires. But the
celebration still carries with it the echoes of an incantation.
Poetry has grown out of magic.
‘Bright star, would I were stedfast as thou art!’ Den lieb’ ich, Her
Unmogliches begehrt. Why do poets crave for the impossible?
Because that is the essential function of poetry, which it has
inherited from magic. In the wild transport of the mimetic
dance the hungry, frightened savages express their weakness in
the face of nature by a hysterical act of extreme mental and
physical intensity, in which they lose consciousness of the external
world, the world as it really is, and plunge into the subconscious,
the inner world of fantasy, the world as they long for it to be. By a
supreme effort of will they strive to impose illusion on reality.
In this they fail, but the effort is not wasted. Thereby the
psychical conflict between them and their environment is
resolved. Equilibrium is restored. And so, when they return to
reality, they are actually more fit to grapple with it than they
were before.
14 Jackson 170.
444 STUDIES IN ANCIENT GREEK SOCIETY XIV
Keats was twenty-four, on his way to Italy in a last effort
to recover his health. He had seen Fanny Brawne for the last
time. Down the Channel his ship was driven by bad weather into
Lulworth Cove, where he went ashore — his last walk on English
soil. He returned to the ship in the evening, and it was then he
composed this sonnet and wrote it outin a copy of Shakespeare’s
poems. Four months later he died in Italy of consumption.
Bright star, would I were stedfast as thou art!
That is a conscious wish — the wish of a dying man. But already
it is charged with poetical memories:
But I am constant as the northern star
Of whose true-fix'd and resting quality
There is no fellow in the firmament.
This sets his own fantasy in motion. He identifies himself,
with the star, and then with the moon, which, as we saw in an
earlier chapter, has been worshipped from the earliest times as
a symbol of everlasting life. And from the moon, still faintly
conscious of the ship rocking gently in the swell running into
the Cove, he looks down on the movement of the tides creeping
to and fro across the contours of this planet:
Not in lone splendour hung aloft the night,
And watching with eternal lids apart,
Like nature’s patient, sleepless Eremite,
The moving waters at their priestlike task
Of pure ablution round earth's human shores,
Or gazing on the soft new-fallen mask
Of snow upon the mountains and the moors —
Then, having withdrawn thus into infinity, still responsive to the
drowsy swaying of the ship, he descends, immortalised, to earth:
No, yet still stedfast, still unchangeable,
Pillow'd upon my fair love’s ripening breast,
To feel for ever its soft fall and swell,
Awake for ever in a sweet unrest,
Still, still to hear her tender-taken breath,
And so live ever —
But it is impossible. There could be no love without death, and
so his prayer for immortality turns into its opposite:
And so live ever, or else swoon to death.
XIV ' THE ART OF POETRY 445
He -wakes up. It is like a dream stirred by the rocking of the
boat — a dream in which the white breast of his sleeping love
is symbolised in the moving waters and the snow on the
mountains. But through the dream he has thrown off what
was oppressing him. He has recovered his peace of mind. The
world is still objectively the same — the world
Where youth grows pale, and spectre-thin, and dies —
but his subjective attitude to it has changed. And so for him
it is not the same. That is the dialectics of poetry, as of magic.
2. Rhythm and Labour
Rhythm may be defined in its broadest sense as a series of
sounds arranged in regular sequences of pitch and time. Its
ultimate origin is of course physiological, but at that level it is
something that man shares with the animals. We are not con-
cerned here with the physical basis of rhythm, but with what
man has made of it. I am going to argue that human rhythm
originated from the use of tools.
We all know that, when children are learning to write, they
often roll the tongue in time with the hand, or even pronounce
the words aloud — not because there is anyone to listen but to
help the fingers guide the pen. What actually happens is that
there is a ‘spread* from the motor organs of the hand to the
adjacent area of the brain, which controls the tongue. As the.
child improves with practice, the spread is eliminated.
Similarly, when a man is doing heavy work, such as lifting a
log or stone, he pauses before the height of each muscular
effort for an intake of breath, which he holds by closing the
glottis; then, as he relaxes after the effort, the glottis is forced
open by the pent-up air, causing a vibration of the vocal
chords — an inarticulate grunt.
Savages, like children, gesticulate when they talk. The
function of gesticulation is not merely to help others to under-
stand. They gesticulate just as much when talking to themselves.
It is instinctive, like the other movements just described. The
movement of the vocal organs overlaps, as it were, with the
other muscular movements. For us, speech is primary,
446 STUDIES IN ANCIENT GREEK SOCIETY XIV
gesticulation secondary, but it does not follow that this was
so with out earliest ancestors. The inherent interdependence of
speech and gesture in primitive psychology is an attested fact . 16
On the strength of these considerations it was argued half a
century ago by Bucher that speech evolved from reflex actions
of the vocal organs incidental to the muscular efforts involved
in the use of tools . 18 As the hands became more finely articulated,
so did the vocal organs, until the awakening consciousness
seized on these reflex actions and elaborated them into a
socially recognised system of communication.
In working out his hypothesis Bucher made an extensive
study of labour songs. The function of these songs is to ex-
pedite the labour of production by imparting to it a rhythmical,
hypnotic character. The spinner sings in the belief that her
song will help the spinning-wheel to go round, and since it
helps her to turn it, it does help the spinning-wheel to go
round. This is very near to magic. In particular instances it can
be shown that these songs originated as incantations . 17
Labour songs abound at
all stages of culture all over
the world — except where
they have been silenced by
the hum of machinery . They
spring spontaneously to the
lips of savages whenever
they are engaged in manual
work, providing, especially
FIG. 65. Balingjo music: Bccotian terracotta
among the women, an irrepressible continuo to the routine
of dally life . 18 And they have a special importance for the
16 Gray FL 155, R. B. Smyth 2. 412, Rattray A 247. Many savage
peoples have elaborate 'deaf-and-dumb' languages which they use to
circumvent taboos of silence: Spencer A 433, 600-8, Howitt NTSEA
(1904) 723-35, R. B. Smyth 2. 4, 308.
18 Bucher 395. Cf. Cic. ZD. 2. 23. 56 profundenda voce omne corpus
intenditur vcnitque plaga vehementior. Since Bucher a somewhat similar
hypothesis, worked out more fully on the physiological side but showing less
insight into the other aspects of the problem, has been advanced by Paget.
17 Chadwick GL 3. 783.
18 Bucher 63-243, cf. Chadwick GL 3. 583, 648, 783, Schapera 285,
Ordc Brown 167, Krige 338, Junod LSAT 2. 207-9.
XIV
THE ART OF POETRY
447
origin of poetry, because in them, with certain significant
modifications, die original relationship between language and
labour has been preserved. This was perceived by Bucher,
whose main conclusion, that the rhythm of human speech is
derived from the labour process, is undoubtedly sound. He
attempted to support it by identifying particular rhythms with
particular processes. 10 This part of his argument was mistaken,
and I have abandoned it. The clearest proof of his conclusion
lies in an analysis of the principles of song structure, and for
this I am responsible.
The work of rowing a boat involves a simple muscular
operation, repeated at regular intervals without variation. The
time is marked for the oarsmen by a repeated cry, which in its
simplest form is disyllabic: O — op! The second syllable marks
the moment of exertion; the first is a preparatory signal.
Hauling a boat is heavier work than rowing. The moments
of exertion are more intense and so are spaced at longer intervals.
This leaves room for expansion of the preparatory syllable, as in
the Irish hauling cry: Ho — li — ho — hup! Sometimes the cry ends
with a syllable of relaxation, as in the Russian hauling cry:
E — uch — nyern! And in many cases it has become partly or wholly
articulate: Heave — o — ho! Haul away!
These two elements, variable and constant, which constitute
the simple, disyllabic labour cry, can be recognised in the
arsis and thesis of prosody, which denote properly the raising
and lowering of the hand or foot in the dance. 20 And so the
ictus or beat of rhythm is rooted in the primitive labour
process — the successive pulls at the log, or the strokes of the
tool on stick or stone. It goes back to the very beginning of
human life, to the moment when man became man. That is
why it stirs us so deeply.
The following ditty was recorded by Junod, the Swiss mis-
sionary mentioned above (p. 439), from a Thonga boy, who
sang it extempore at the roadside while breaking stones for his
European employers:
J Ba hi shani-sa, ch£l
Ba ku hi hlupha, ch£l
Ba nwa xnakhofi, ehil
Ba nga hi nyiki, eh£l
10 Bucher 407. 20 lb. 25, 402.
448 STUDIES IN ANCIENT GREEK SOCIETY XIV
They treat us badly, eh&
They are hard on us, eh£l
They drink their coffee, eh&
And give ns none, eh£l®i
The repeated ell! is the labour cry, marking the hammer-
strokes. This is the constant. Each time it is prefaced with a
few articulate words improvised to express the worker's sub-
jective attitude to his task. The song has grown out of the
cry, just as the cry has grown out of the work itself.
Heave on, cut deepl
How leaps my fluttering heart
At the gleam -that flashes from thine eyes,
O Puhi-huial
Heave on, cut deep! 22
That is a Maori rowing song. The boatswain uses the cries
intermittently, and between them he improvises a compliment
to the chief's daughter travelling in the boat. During the im-
provisation the time is marked by the rhythm of the words.
The cry is still functional, but it is on the way to becoming a
refrain.
My next example is the Volga Boat Song:
E-uch-nyem! e-uch-nyem! Yeshcho razikl yeshcho da raz!
Razovyom my beryozu, razovyom my kudryavu!
Aida da, aida! razovyom! aida da, aida! kudryavu!
E-uch-nyem! e-uch-nyem! Yeshcho razik! yeshcho da raz! 23
Here an improvised exhortation to the task is prefaced and
concluded with the hauling cry, which contains it and defines
it.
The labour song was developed by expanding the improvised
variable between the moments of exertion. The workers ran
over dreamily scraps of traditional lore or passed desultory
comments on current events — whatever was uppermost in
their minds. We possess an ancient Greek milling song —
'Grind, mill, grind’ — interspersed with allusions to the tyrant
Pittakos ; 24 and there is another with the same refrain in modern
si Junod LSAT 2. 284. 22 Andersen 373.
23 Bucher 235. There are many versions, because the middle of the stanza
is still improvised.
24 Cam. Pop. 30.
XIV ■ THE ART. OF POETRY 449
Greek, improvised by a woman -forced to grind barley for a
police squad searching for her husband . 88 The constant, tied
to the task in hand, tends to remain unchanged; the variable
varies indefinitely from day to day. Many of die obscurities in
our folk-songs probably arise from our ignorance of the cir-
cumstances -that inspired the particular form in which they
survive. Other examples of the same type will be found among
the negro spirituals, which inculcate Bible teaching at the
cqrrw> time as they soothe the labourers at their task , 88 and in
the English' sea-shanties, like this one from the end of the
eighteenth century:
Louis was the King of France afore the Revolution,
Away, haul away, boys, haul away together!
Louis had his head cut off, which spoilt his constitution,
Away, haul away, boys, haul away together ! 87
Meanwhile the art of song had broken away from the
labour process. Songs were improvised at leisure, when the
body was at rest. But they conformed to the traditional
pattern. This is from Central Africa, where it was sung one
evening round the camp fire by the porters attached to a white
man's caravan:
The wicked white man goes from the shore— puti, putif
We will follow the wicked white man — puti, puti!
As long as he gives us food — puti, puti!
We will cross the hills and streams — puti, puti!
With this great merchant’s caravan — puti, pud! 28
And so on, one after another round the fire, till they all fell
asleep. The improvisations were rendered in turn by in-
dividuals, while the repeated puti (which is said to mean
‘grub’) was sung by all in unison. This gives us the familiar,
universal structure of solo and chorus . 89 The labour cry is now
nothing but a refrain.
Severed from the labour process, the constant too is ex-
panded. It becomes fully articulate, and is varied so as to
diversify the rhythmical pattern, but without destroying the
85 Polites E no. 234. as Biicher 263-73.
87 Jh 239. as Burton 361-2. .
89 Basedow 376, Codrington 335, Layard 315, 6u, Orde Brown 167,
RA. Talbot 808, Driberg 127, 129, 245, Chadwick GL 3. 353, 355-6,
581, Entwhistle 19, 35. ? ?
Eb
45 ° STUDIES IN ANCIENT GREEK SOCIETY XIV
sense of regular repetition, on which the unity of the whole
depends.
Why does your brand sae drop wi’ blude,
Edward, Edward?
Why does your brand sae drop wi* blude,
And why sae sad gang ye, O?
O, I hae kill’d my hawk sae gude,
Mither, mither,
O, I hae kill’d my hawk sae gude,
And I had nae mair but he, O.
And so we reach the ballad quatrain, in which the refrain has
disappeared as such, but is still embedded in the rhythmical
structure, which rests on a continual alternation of thesis and
antithesis, announcement and responsion:
There liv’d a lass in yonder dale,
And down in yonder glen O,
And Kathrine Jafiray was her name,
Well known by many men O. 3 o
In the ballad measure, the stanza is a musical 'sentence', the
couplet a musical 'phrase', the verse a musical 'figure', There
are two figures in each phrase, two phrases in each sentence.
The members of each pair are complementary, similar yet dif-
ferent. This is what musicologists call binary form: AB.
This musical anatomy of the ballad measure is not merely an
analogy. It is the only proper method of analysis. The prosody of
our textbooks is as remote from the living history of poetry, as
conventional grammar is from the living history of language.
The ballad was originally a dance, as it still is in some parts of
Europe, like this one from the Faroe Islands:
The precentor sings the ballad and the rhythm is stamped with the feet.
The dancers pay close attention to his words, which must come clearly, since
the characteristics of the narrative are brought out by the mime. Hands are
tightly clasped in the turmoil of battle; a jubilant leap expresses victory.
All the dancers join in the chorus at the end of each stanza, but the stanza
itself is sung only by one or two persons of special repute. 31
The analytical principles of musicology belong to the study of
rhythm as such, that is, to the common foundation of poetry,
music and dancing.
33 Gummere 169, 263.
31 Entwhisdc 35.
XIV
THE ART OF POETRY
451
Most of our folk-songs are in binary form, but some are
more elaborate. In the- Volga Boat Song, for instance, the
stanza consists of an improvised passage preceded and followed
by the verse containing the traditional hauling ciy. In musical
terminology, the first subject is followed by a second, and then
the first is repeated or resumed. This is ternaty form: ABA. In
skilful hands A2 becomes something more than a repetition of
Ai: it is Ai in a new form conditioned by B. And so ternary
form is more organic, more dialectical than binary. That is
why it has been so highly cultivated in modern music . 88 Both
forms were used by the Greeks. The melody of Greek music
has perished; but since most of their poetry, apart from epic
and dramatic dialogue, was composed for singing, its rhythm
can be recovered from the words. I have discussed this in
my Greek Lyric Metre, where the Greek strophe is shown to be
an organism of exactly the same type as die modern stanza.
We shall return to this subject in the next chapter.
To resume. The three arts of dancing, music and poetry
began as one. Their source was the rhythmical movement of
human bodies engaged in collective labour. This movement had
two components, corporal and oral. The first was the germ of
dancing, the second of language. Starting from inarticulate
cries marking the rhythm, language was differentiated into
poetical speech and common speech. Discarded by the voice
and reproduced by percussion with the tools, the inarticulate
cries became the nucleus of instrumental music.
The first step towards poetry properly so called was the
elimination of the dance. This gives us song. In song, the
poetry is the content of the music, the music is the form of the
poetry. Then these two diverged. The form of poetry unac-
companied by music is its rhythmical structure, which it has
inherited from song but simplified so as to develop its logical
content. Poetry tells a story, which has an internal coherence
of its own, independent of its rhythmical form. And so later
there emerged out of narrative poetry the prose romance and
novel, in which poetical speech has been replaced by common
speech and the rhythmical integument has been shed — except
that the story itself is cast in a balanced, harmonious form.
88 Macpherson 61-90.
45^ STUDIES IN ANCIENT GREEK SOCIETY XIV
Meanwhile there has grown up a type of music which is
purely instrumental. The symphony is the antithesis of the
novel. If the novel is speech without rhythm, the symphony is
rhythm without speech. The novel derives its unity from the
story it tells, taken from perceptual life; the symphony draws
its material entirely from fantasy. It has no internal coherence
apart from its form. Hence all those structural principles
which have disappeared in the novel have been elaborated in
music to an unprecedented degree. They have come to be
regarded as the special province of music. We speak of them
habitually as ‘musical form*. Yet they can still be traced in
poetry — in the arrangement of its subject matter, I mean, not
merely in its metrical structure — if we study it with a sense of
music. Let us examine two examples, which, besides illustrat-
ing the point at issue, will show once again how poetry is
related to magic.
Sappho’s Ode to Aphrodite is the oldest European lyric; and*
it is a lyric in the full sense — a song sung to the lyre. Sappho
was head of a religious society of young ladies, dedicated to.
Aphrodite. One of these girls, to whom she is passionately
devoted, has failed to reciprocate her love.
Aphrodite, goddess enthroned in splendour,
Child of Zeus Almighty, immortal, artful,
I beseech thee, break not my heart, O Queen, with sorrow and anguish!
Rather come, O come as I often saw thee,
Quick to hear my voice from afar, descending
From thy Father’s mansion to mount thy golden chariot drawn by
Wings of sparrows fluttering down from heaven
Through the cloudless blue; and a smile was shining,
Blessed Lady, on thy immortal lips, as standing beside me
Thou didst ask: ‘Well, what is it now? what is that
Frantic heart’s desire? Do you need my magic?
Whom then must I lure to your arms? Who is it, Sappho, that wrongs you?
On she flies, yet soon she shall follow after;
Gifts she spurns, yet soon she shall be the giver;
Love she will not, yet, if it be your will, then surely she shall love’.
So come now, and free me from grief and trouble!
Bring it all to pass as my heart desires it!
Answer, come, and stand at my side in arms, O Queen, to defend me!
XIV THE ART OF POETRY 453
Sappho begins by stating her prayer. She goes on to recall
how similar prayers had previously been answered. And
then the prayer is repeated. This is ternary form, treated
dynamically by a conscious artist. The prayer opens negatively,
tentatively; it ends positively, confidently, as though, thanks
to what has come in between, a favourable answer were assured.
What does come in between? She reminds Aphrodite of the
past: ‘If ever before ... so now’. That was traditional. When
you prayed to the gods, you reinforced your appeal by reminding
diem of previous occasions when you had received their help
or earned their gratitude .® 3 It was a ritual formula. And ritual
takes us back to magic. In magic you enact in fantasy the ful-
filment of the desired reality. And that is what Sappho does
here, except that there is no action, no dancing, only a flight of
the imagination. She beseeches the goddess to come; then en-
visages her as coming — sees her, hears her; and then, inspired
by this imaginative effort to greater confidence, she repeats
her prayer. It is magic transmuted into art.
In English poetry, being less close to music, such survivals
of musical form are only sporadic, and so the literary critics,
who are not interested in origins, have failed to notice them.
And yet they are all familiar with this sonnet of Shakespeare's,
which is as perfect an example of ternary form as any to be
found in Greek:
33 II.
When, in disgrace with fortune and men’s eyes,
I all alone beweep my outcast state,
And trouble deaf heaven with my bootless cries,
And look upon myself, and curse my fate,
Wishing me like to one more rich in hope,
Featur’d like him, like him with friends possest,
Desiring this man’s art and that man’s scope.
With what I most enjoy contented least,
Yet in these thoughts myself almost despising,
Haply I think on thee, and then my state,
Like to the lark at break of day arising
From sullen earth, sings hymns at heaven's gate;
For thy sweet love rememb’red such wealth brings
That then I scorn to change my state with kings.
1. 39-42, 394—5,453, 5. 116-7, 16. 236-8, Oi. 4. 763-6, Pi. O. X.
454 STUDIES IN ANCIENT GREEK SOCIETY XIV
Only one critic has explained the structure of this poem, and
he was a musicologist. 34 In fourteen lines the poet revolu-
tionises his attitude to the world. At the beginning he is an
outcast, crying to deaf heaven; at the end a king, singing
hymns at heaven’s gate. And the revolution turns on the word
state. At first it connotes despair — the minor key; but when it
returns its tone is modulated, and so we are carried forward to
the ringing triumph of the close.
A revolution in our attitude to the world. Arguing from the
content of poetry — incantations, seasonal songs, that sonnet of
Keats — we concluded that this was the essential function of
poetry. The same conclusion has now been reached from our
study of its form.
3. Improvisation and Inspiration
With us poetry is seldom, if ever, improvised. It is .a matter
of pen and paper. There must be many contemporary poets
whose melodies are literally unheard. They have been written
down by the poet, printed, published, and read in silence by
individual purchasers. Our poetry is a written art, more
difficult than common speech, demanding a higher degree of
conscious deliberation.
It is important to remember that this feature of modern
poetry is purely modern. In antiquity and the middle ages,
and even to-day among the peasantry, the poet is not divided
from his audience by the barrier of literacy. His language is
different from common speech, but it is a spoken language,
common to him and his audience. He is more fluent in it than
they are, but that is only because he is more practised. To some
extent they are all poets. 86 Hence the anonymity of most
popular poetry. Generated spontaneously out of daily life, it
passes, always changing, from mouth to mouth, from parents
34 Hadow 10-2, cf. G. Thomson AO 1. 14.
36 Chadwick GL 3. 65, 178, 659, Layard 3x4-5, Schapera 285. On
improvisation see Chadwick GL. x. 578, 3. 64-5, 152, 156, 174, 181-3,
187, 213, 412, 529, 583, 616, 647-8, 659-63, 868, Jeanroy 357, Schapera
405, Driberg 129, Junod BH 85, Orde Brown 167, Bonwick 29, W.
Bateson 165-6, Layard 3 14—5, P. A. Talbot 808. Greek drama originated
in improvisation (Arist. Pcit. 4. 14): see below p. 467.
XIV
THE ART OF POETRY
455
to children, until the faculty of improvisation decays. Only
then does it become 'fixed, and even then it preserves a dis-
tinctive quality, which we describe by saying that, however
perfect it may be in point of craftsmanship, it lacks the quality
of conscious art. That is just what it does lack — the stamp of an
individual personality; and inevitably so, because it is the
product not of an individual but of a community. The primi-
tive poet is not conscious of his medium as something different
from common speech; and in fact, as we have seen, the dif-
ference is less. Hence he is able to improvise. As he succeeds
in objectifying his medium, he loses the gift of improvisation,
but at the same time acquires the power of adapting it to his
own personality, and so becomes a conscious artist.
On the other hand, the effect of poetry is still, as it has always
been, to withdraw the consciousness from the perceptual
world into the world of fantasy. In comparing poetical speech
with common speech we saw that it was more rhythmical,
fantastic, hypnotic, magical. Now, in our conscious life, all
the factors that make up our distinctive humanity — economic,
social, cultural — are fully active: individual differences are at
their maximum. Hence, just as the mental processes of con-
scious life reveal the greatest diversity between individuals, so
common speech, which is their medium, is marked by the
greatest freedom of individual expression. But when we fall
asleep and dream, withdrawing from the perceptual world, our
individuality becomes dormant, giving free play to those basic
impulses and aspirations, common to all of us, which in con-
scious life are socially inhibited. Our dream world is less in-
dividualised, more uniform than waking life.
Poetiy is a sort of dream world. Let me quote from Yeats:
The purpose of rhythm is to prolong the moment of contemplation, the
moment when we are both asleep and awake, which is the one moment of
creation, by hushing us with an alluring monotony, while it holds us waking
by variety, to keep us in that state of perhaps real trance, in which the mind
liberated from pressure of the will is unfolded in symbols. 8 ®
One might quarrel with the word ‘liberated’, but that does
not matter now. The language of poetry, being rhythmical, is
hypnotic. Not so hypnotic as to send us to sleep altogether. If
8 ® Yeats 195-6.
45 ^ STUDIES IN ANCIENT GJR E E K SOCIETY- XIV
we analyse any metre in any language, we find in it precisely
that combination of monotony and variety, that interplay
of like and unlike, which, as Yeats perceived, is needed to
hold the mind suspended in a sort of trance, the special spell
of poetry, caught between sleep and waking in the realm of
fantasy.
And so, when we say a poet is inspired, we mean that he is
more at home than other men in this subconscious world of
fantasy. He is exceptionally prone to psychical dissociation.
And through this process his inner conflicts — the contradictions
in his relationship to society — are discharged, relieved. The
discords of reality are resolved in fantasy. But, since this
world into which he retires is common to him and his fellow
men, the poetry in which he formulates his experience of
it evokes a general response, expressing what his fellows feel
but cannot express for themselves, and so draws them all into
a closer communion of imaginative sympathy:
Und wenn der Mensch In seiner Qual verstummt,
Gab mir ein Gott, zu sagen, wie ich Ieide. 8 ?
His fellows are tormented by unsatisfied longings which they
cannot explain, cannot express. He too 'is unable to explain
them, but thanks to the gift of inspiration he can at least
express them. And when he expresses them they recognise his
longings as their own. As they listen to his poetry they go
through the same experience as he did in composing it. They
are transported into die same world of fantasy, where they find
the same release.
In the mimetic dance, directed by their leader, the savage
huntsmen pre-enact the successful prosecution of the hunt,
striving by an effort of will to impose illusion on reality. In
fact all they do is to express their weakness in the face of
nature, but by expressing it they succeed to some extent in
overcoming it. When the dance is over, they are actually better
huntsmen than they were before. In poetry we see the same
process at a higher level. Civilised man has succeeded to a
large extent in mastering nature, but only by complicating
his social relations. Primitive society was simple, classless,
Goethe Tasso 3432-3.
XIV
THE' ART OF POETRY
457
presenting a weak but united front against nature. Civilised
society is more complex, richer, more powerful, but, as a
necessary condition of all this, it has always hitherto been
divided against itself. Hence the conflict between society and
nature— the basis of magic — is overlaid by a conflict between
the individual and society — the basis of poetry. The poet does
for us what the dance-leader does for his fellow savages.
The primitive poet does not work alone. His audience col-
laborates. Without the stimulus of a listening crowd he cannot
work at all. He does not write, he recites. He does not com-
pose, he improvises. As the inspiration comes to him, it pro-
duces in his listeners an immediate response. They surrender
to the illusion immediately and wholeheartedly. In these cir-
cumstances the making of poetry is a collective social act.
When we read a poem, or hear one being read, we may be
deeply moved, but we are seldom completely ‘carried away'.
The reaction of a primitive audience is less sublimated. The
whole company throw themselves into the world of make-
believe: they forget themselves. I have seen this many times in
the west of Ireland. Or listen to this account of a Russian
minstrel reciting in a hut on one of the islands on Lake
Onega:
Utka coughed. Everybody became silent. He threw his head back and
glanced round with a smile. Seeing their impatient, eager looks, he at once
began to sing. Slowly the face of the old singer changed. All its cunning
disappeared. It became childlike, naive. Something inspired appeared in it.
The dovelike eyes opened wide and began to shine. Two litde tears sparkled
in them; a flush overspread the swarthiness of his cheeks; his nervous throat
twitched. He grieved with Ilya of Murom as he sat paralysed for thirty
years, gloried with him in his triumph over Solovey the robber. AH present
lived with the hero of the ballad too. At times a cry of wonder escaped from
one of them, or another’s laughter rang through the room. From another
fell tears, which he brushed involuntarily from his lashes. They all sat
without winking an eye while the singing lasted. Every note of this monoton-
ous but wonderfully gentle tune they loved. 3 ®
These people were illiterate; yet poetry meant something for
them which it certainly does not mean for English people
to-day. We have produced Shakespeare and Keats, it is true,
and they were greater than Utka. But Utka was popular, and
as Quoted by Chadwick GL z. 241.
45 $ STUDIES IN ANCIENT GREEK SOCIETY XIV
that is more than can be said of Shakespeare or Keats in our
country to-day.
Let us push on from Russia into Central Asia and see how,
sixty years ago, the Turkmens listened to their poetry:
When I was in Etrek, one of these minstrels had a tent dose to ours, and
as he visited us of an evening, bringing his instrument with him, there
flocked around him the young men of the vicinity, whom he was constrained
to entertain with his heroic lays. His singing consisted of forced guttural
sounds, more like a rattle than a song, and accompanied at first with gentle
touches on the strings. But as he became excited the strokes grew wilder.
The hotter the batde, the fiercer the ardour of the singer and his youthful
listeners; and really the scene assumed the appearance of a romance, as the
young nomads, uttering deep groans, hurled their caps into the air and
dashed their hands in a passion through their hair, as though they were
furious to combat with themselves. 89
These Turkmens, poet and listeners alike, were literally
entranced.
Turning to ancient times, we may recall a Byzantine
writer’s visit to the court of Attila:
When dusk fell, torches were lit, and two Huns came out in front of
Attila and chanted songs in honour of his victories and martial prowess. The
banqueters fixed their eyes on the singers, some of them enraptured, others
greatly excited as they recalled the fighting, while those whom old age had
condemned to inactivity were reduced to tears. 49
This is the context in which we must study the Iliad and
Odyssey. How did the ancient Greeks react to Homer? We are
apt to assume that they behaved just like ourselves, but this is a
mistake. In one of Plato’s dialogues a Homeric minstrel
describes the effect of his recitals on himself and his audience:
When I am narrating something pitiful, my eyes fill with tears; when
something terrible or strange, my hair stands on end and my heart throbs. . . .
And whenever I glance down from the platform at the audience, I see them
weeping, with a wild look in their eyes, lost in rapture at the words they
hear. 41
When such poets are questioned about the nature of their
art, they all give the same answer. They all claim to be in-
spired in the literal sense of the word — filled with the breath of
God:
39 Vambeiy 322. 49 Prise. Z=FHG. 4. 92. 41 PI. Io 535.
XIV
THE ART OF POETRY
459
A skilled minstrel of the Kirghiz can recite any theme he wants, any story
that is desired, extempore, provided only that the course of events is clear to
him. When I asked one of their most accomplished minstrels whether he
could sing this or that song, he answered: 'I can sing any song whatever, for
God has implanted this gift of song in my heart. He gives the words on my
tongue without my having to seek diem. I have learnt none of my songs. All
springs from my inner self,’ 42
We remember Caedmon, inspired by an angel that visited him
in dreams , 43 and Hesiod, who was taught by the Muses while tend-
ing his flocks oh Helikon , 44 and Phemios and Demodokos, the
minstrels of Ithaca and Phseacia: ‘I am self-taught’, says Phemios,
‘for God has implanted all manner of songs in my heart ’. 46
For primitive peoples everywhere the poet is a prophet, who
being inspired or possessed by a god speaks with the god’s
voice. For the ancient Greeks the connection between prophecy
(mantikf) and madness (mania) was apparent in the words them-
selves. To them the magical origin of poetry and prophecy was
self-evident, because the symptoms of both reminded them of
the orgiastic dances that survived in their cults of Dionysus:
All good epic poets are able to compose not by art but because they are
divinely inspired or possessed. It is the same with lyric poets. When com-
posing they are no more sane than the Korybantes when they dance. As soon
as they engage in rhythm and concord, they become distracted and pos-
sessed, like the Bacchants who in their madness draw milk and honey from
the streams. 46
These religious devotees were subject under the influence of
music to hysterical seizures, which were explained by saying
that they were (nthtoi, that they had ‘a god in them’. 4 ’ At this
level we can no longer speak of art. We have reached its roots
in magic.
Inspiration and possession are the same thing. In primitive
society mental disorders involving loss of consciousness and
convulsions are attributed to possession by a god or animal or
ancestral spirit . 48 This idea emanates from the ecstasy of the
42 Radlov PV 5. xvii. 43 Bede Zccl. Hist. 4. 24. 44 Hes. Th. 22-3.
46 Oi. 22. 347-8, 8. 479-81. 46 PL lo 533e.
4 ’ G. Thomson AA 374 . 377-8.
48 Junod LSAT 2. 479-503, Smith 6c Dale 2. 136-52, SchaperaBTSA
253, Roscoe B (1911) 2,74, 318, 320-2, Codrington.218, Chadwick GL
3. 449, 454, Czapliclca 307-25, Karsten 18, Earthy 199, Webster 151,
175, Fallaize in Hastings xo. 122.
460 STUDIES IN ANCIENT GREEK SOCIETY XIV
mimetic dance, in which the performers lose consciousness of
their identity as they impersonate the totemic animal — the
symbol of the heightened common ego evoked by the dance.
Hysteria is a neurosis — a conflict between the individual
and his environment which issues in a revolt of the subcon-
scious. It is common among savages, not because they are more
prone to such conflicts than we are, but because their conscious-
ness is shallower, less resilient. It is treated by magic. When
the first symptoms appear, a song is chanted over the patient.
This facilitates the psychical dissociation, precipitates the fit . 48
Here, then, we have poetry at a purely magical level, or rather
not poetry at all but the form of therapeutic magic out of
which poetry evolved. For magic too is a revolt of the sub-
conscious, cured in the same way. The difference is that in the
mimetic dance this hysterical propensity is organised collec-
tively — it is organised mass hysteria; whereas these individual
seizures are sporadic. But the treatment is essentially the same.
The patient is exorcised. The possessing spirit is evoked and
expelled by the magic of the song. The exorcist who administers
the treatment — the shaman, medicine-man or witch-doctor, as
he is variously called — is usually himself a hysterical subject
who has undergone a special training . 80 The relation of the
exorcist to the patient is thus similar to that of the leader to
his followers in the mimetic dance.
Prophecy is a development of possession. One of the com-
monest conditions of exorcising a patient is that the possessing
spirit should be forced to reveal its name, and often, after
revealing its name, it demands to be propitiated in return for
releasing its victim. In this way the procedure becomes a
means of proclaiming the will of the gods and so of predicting
the future. The hysterical seizure assumes the form of a pro-
phetic trance, in which the patient becomes a medium in the
modern spiritualistic sense — a vehicle for the voice of a god
or spirit . 81 In this condition he expresses fears, hopes, anti-
cipations of the future, of which in his conscious life he is
unaware. We still say that coming events cast their shadows
before. They impingeonour subconscious, causing an indefinable
48 Fallaize l.c., Smith & Dale 2. 137-8.
80 G. Thomson AA 375. 81 lb. 376.
XIV THE ART OF POETRY 461
unrest, and in the prophet, whose subconscious, being abnor-
mally active, is constantly liable to erupt, they rise to the
surface.
And finally the prophet becomes a poet. In primitive thought
there is no clear line between prophecy and poetry. The
minstrels described in the Homeric poems are credited with
second sight, and their persons are sacrosanct . 62 The poet is
the prophet at a higher level of sublimation. The physical
intensity of the trance has been mitigated, but it is a trance all
the same. His psyche is precipitated into fantasy, in which his
subconscious struggles and aspirations find an outlet. And just
as the prophet's predictions command general acceptance, so
the poet's utterance stirs all hearts.
In this way we are able, with Caudwell, to define the essential
nature of art:
Art changes the emotional content of man's consciousness so that he can
react more subtly and deeply to the world. This penetration of inner reality,
because it is achieved by men in association and has a complexity beyond the
power of one man to achieve, also exposes the hearts of his fellow men and
raises the whole communal feeling of society to a new plane of complexity. It
makes possible new levels of conscious sympathy, understanding and
affection between men, matching the new levels of material organisation
achieved by economic production. Just as in the rhythmic introversion of
the tribal dance each performer retired into his own heart, into the fountain
of his instincts, to share with his fellows not a perceptual world but a world
of instinct and blood-warm rhythm, so to-day the instinctive ego of art is
the common man into which we retire to establish contact with our fellows. 68
There is one other aspect of inspiration that may be men-
tioned here. Just as magic was for a long time the special
province of women, so we find all over the world that inspira-
tion in prophecy and poetry belongs especially to them . 64
The evidence is all the more striking because their part in
primitive life is not nearly so well documented as the men’s . 66
62 Hes. 'll. 31-2, Oi. 8. 479-81, 22. 345-6.
68 Caudwell 155.
64 Bucher 434-52, Briffault 2. 5x4-71, Chadwick GL 3. 186-8, 413,
663, 895-8. Of 1202 songs collected in Estonia, Latvia and Lithuania
678 are women’s songs, 355 are men's, and 169 are indeterminate (Bucher
450). The history of ballad poetry in southern and western Europe points
to a similar conclusion: Entwhistle 37-8.
66 See p. 241 and cf. Bucher 435-6, Chadwick GL 3. xxii.
462 STUDIES IN ANCIENT GREEK SOCIETY XIV
I am not going to enlarge on this subject now. The reader may
study it in the pages of Bucher, Briffault, and Chadwick. It was
more than a poet's fancy that prompted Homer and Hesiod to
invoke the aid of female deities. The woman’s part in the
origin of music is commemorated in the word itself.
XV
THE RITUAL ORIGINS OF GREEK EPIC
i. The Problem
Art grows out of ritual. Stated in general terms, that is a pro-
position no serious student would deny. Many, it is true, dis-
miss it as irrelevant, but that is because, like magic, it is or can
be a great force in the world, and they want to keep it tame.
By despising the study of its origins they curtail their power
to understand it and so to enjoy it. We must rescue it from
them. They have no right to clip its wings.
The ritual ties of Greek poetry are for the most part self-
evident. The Homeric Hymns, the Pindaric Odes, Attic
drama, were all conscious acts of worship. Only in epic are
they not apparent. The historical criticism of Greek epic has
been directed along two .main channels. For a hundred years
and more a host of classical scholars have been debating the
anatomy of the Iliad and Odyssey. The controversy is not yet
settled, but recently it has shown signs of flagging from sheer
exhaustion. Meanwhile a new lead has been given by a pro-
fessor of Anglo-Saxon. Applying the comparative method to the
epics of different peoples, Chadwick has established a number
of correlations which make it possible to refer this kind of
poetry to a specific set of social and historical conditions. But
the problem of its ritual origins remains.
The three main forms of Greek poetry were, in the order of
their maturing, epic, lyric, and drama. The Iliad and Odyssey , in
the main, can hardly be later than the eighth century,* Alkman,
our earliest survivor from the wreckage of lyric, belongs to the
seventh; ^Eschylus makes his dihut at the beginning of the fifth.
This is the chronological order, but all it tells us is when each
form reached the level of conscious art. If we look at them from
the standpoint of their origins, the chronological order is
reversed. Drama combines song, dance, and impersonation; it
preserves the original unity of mimetic magic. Choral lyric
XV
464 STUDIES IN ANCIENT GREEK SOCIETY
combines song and dance. Epic is merely recitation. _Lyric is
based on the strophe or stanza; in epic there is no trace of the
strophe. Thus, the least differentiated of the three, and hence
the most primitive, was the last to mature; the first to mature
was the least primitive. But even this is not the whole truth.
Drama includes recitation, and, though its structure is the most
primitive in the sense of being the oldest, its technique is not,
nor is its content. In these respects it marks the consummation
of all three.
These complications have eluded the empiricists, with the
result that a scientific history of Greek poetry has never been
attempted. Yet they are not difficult to explain. These three
art forms correspond to three successive phases in the growth of
Greek society — the early monarchy, the landed aristocracy,
democracy. Their mutual contradictions fall into place as soon
as it is realised that they reflect the dialectics of the class-
struggle.
The problem of the present chapter will be studied under
three main heads: the structure of the strophe, the evolution
of the chorus, and the relations of the sexes. The reader is
doubtless wondering what bearing these questions can pos-
sibly have on the Iliad and Odyssey. Well, we shall see.
2. The Strophe
Stanza and strophe are one and the same thing. The stanza
is a ‘stand’ or ‘pause’; the strophe is a ‘turn’, like the Latin
versus. Both denote properly divisions in the movement of a
dance.
In English poetry there are two principal types of ballad
measure — the short couplet of eight stresses and the long
couplet of fourteen . 1 In the latter the couplet is commonly
subdivided, yielding the familiar ballad quatrain. In this there
are four verses, with four and three stresses alternately. Its
binary structure is marked by the rhyming, which is confined to
the second and fourth verses, that is, to die end of each phrase.
The rhymes thus coincide with the two pauses, minor and
major, in the dance movement. They are as it were echoes of
1 Gummere 307-9.
XV THE RITUAL ORIGINS OF GREEK EPIC 465
the decisive, final steps in each run of the dancers’ feet. That
is the origin of rhyme. It is derived from the vocal accom-
paniment to a co-ordinated bodily movement.
In the ballad quatrain the rhythmical structure has been
reduced to the smallest compass compatible with the
preservation of its organic unity. But the Greek strophe
stands much nearer to its choral origin. All Greek lyric — that
is, all poetry composed in strophic form — was accompanied by
the lyre or flute, and, with the exception of the monody, it
was danced by a chorus. Its structure is consequently ampler and
more elaborate, reproducing the intricacies of the musical
accompaniment and the evolutions of the dancers’ hands and
feet.
There are three types of strophic form — monostrophic,
triadic, and antistrophic. In the monostrophic ode a single
strophe is continuously repeated, exactly like the stanza in
modern verse (aaa). The triad consists of a strophe followed
first by an antistrophe, which is simply the strophe repeated,
and then by an epode, a system composed of the same or
similar rhythmical materials but differently arranged and
serving as a coda (aab); and this triad is continuously repeated
(aab aab aab). Antistrophic form is a series of pairs, strophe
and antistrophe. The members of the pair are identical with
one another, but each pair differs from die last (aa bb cc).
The earliest surviving odes — by Alkman (c. 660 b.c.),
Alkaios and Sappho (630-580 b.c.) are all monostrophic.
Many of the odes by Alkaios and Sappho are monodies, sung
by a soloist without dancing. The triad is said to have been
invented by Stesichoros, who belonged to the following
generation. It was always choral, and became the dominant
form of the later aristocratic convention. Nearly all Pindar's
odes are triadic. Antistrophic form — also choral — is confined
to drama. Such is the chronological order. Our problem is to
reconstruct the development.
As soon as we set diem in their historical context, we en-
counter some significant complications. Alkman lived at
Sparta, whose aristocracy was then at the outset of its long
career, but he was a native of Sardeis in Lydia. The Spartan
poetry of this period was mainly the work of foreigners. We
Ff
466 STUDIES IN ANCIENT GREEK SOCIETY XV
hear also of Terpandros from Lesbos' and Thaletas from Crete.
Moreover, a metrical analysis of Alkman reveals affinities
with Alkaios and Sappho so close that all three must be assigned
to a common Graeco-Anatolian tradition.
Alkaios and Sappho were natives of Lesbos. They belonged
to the generation after Terpandros, and they remained at
home. Both were aristocrats, but the Lesbos of their day was
on the verge of a democratic revolution. Remembering this,
we shall not be surprised to find that their work is more
advanced than Alkman’ s.
Stesichoros was born at Himera, a colony in Sicily founded
jointly by Dorians from Syracuse and lonians from Chalkis.
Like Alkman, he used the Doric dialect, but his technique was
different. There is no reason to doubt that he invented the
triad, but of course he did not invent it out of nothing. He was
working on pre-existing material. The structure of the triad
presupposes a chorus divided into two semi-choruses — the two
sexes, two clans, two age-groups, or whatever they may have
been — which chanted the strophe and antistrophe antiphon-
ally and the epode in unison. But, so far as we know, none of
the extant triadic odes was actually antiphonal. They were sung
in unison. The practice of antiphony had been abandoned, but
the structure remained. What Stesichoros did, then, was to
divest this ritual form of its ritual function and establish it as an
art form.
In our third type, antistrophic form, the repetition has been
reduced to a minimum. That is its distinctive feature. It is the
most flexible of the three, and therefore the most dramatic.
Seeing that it is peculiar to drama, we may infer that the
dramatists invented it.
If the epode of the triad was designed to be sung in unison,
it was in origin a refrain. And if it had originally been appended
to the strophe as well as the antistrophe {ax ax ax), we are
back at the primitive binary sequence of solo and chorus, im-
provisation and refrain. There are several reasons for thinking
that this was in fact the case.
To begin with, if we examine antistrophic form, we reach the
same result. In some of their odes the dramatists employ the
epode, but only as a single, final coda marking the conclusion
XV THE RITUAL ORIGINS OF GREEK EPIC 467
of the whole (aa BB- cc d). They also use another type of coda
known as the ephymnion. This is always in some simple,
popular rhythm and it is appended to both members of the
pair (ax ax bx bx). And this arrangement differs in only one
particular — the disparity of the pairs, which we have just
recognised as an innovation — from the- original form of the
triad as suggested above.
Tragedy, according to Aristotle, was descended from the
dithyramb. This was a type of choral ode which is known to
have been performed in early times by a leader and chorus. The
leader delivered a series of improvised stanzas, while the chorus
interpellated the refrains . 2 With this to guide us, the evolu-
tion of antistrophic form becomes plain. It began with the
primitive sequence of solo and refrain. In the second stage, the
soloist disappeared. The whole ode was sung by the chorus —
a monostrophic ode with ephymttia. In the third, the ode was
made more flexible by dividing it into antistrophic pairs. And
finally the cphymnia were discarded, leaving us with the typical
antistrophic ode.
Further, we must remember that the surviving specimens of
Greek lyric are almost all masterpieces of conscious art. The
odes employed in the everyday worship of the temples must
have been less elaborate. We know little about these, but
enough to show that in them the solo-and-chorus convention
survived throughout antiquity. In the Christian liturgy it is
still alive to-day. It survived in the dirge, in which She im-
provisations of the leaders were answered by inarticulate
wails ; 3 we find it again in the Cretan Hymn of the Kouretes , 4
and in the Hymn to Dionysus from Elis. The latter is quoted
by Plutarch, who speaks of its refrain as an epode . 5
Lastly, there is die word itself. What does epoidSs mean? In
reference to the third member of the triad it was explained as an
‘after-song', a coda. But this was a technicality. In popular
language it meant a ‘charm’, ‘spell', or ‘incantation’, a song
‘sung over’ somebody, like the dirge over Hector’s corpse, or
the spell over the sick man to heal him, or the curse over the
2 Archil. 77, cf. Pickard-Cambridge 19. On the Greek solo-and-chorus
convention see H. W. Smyth xxi, xl, xlvi, xlviii, cxi, cxv, cxvi, cxxii, 503.
a II. 24. 719-76. 4 Diehl 2. 279-81. 8 Plu. M. 299b.
468 STUDIES IN ANCIENT GREEK SOCIETY XV
criminal to damn him. This is undoubtedly the primary
meaning of the word. The refrain was originally an incanta-
tion. In the Oresteia the Erinyes perform a magical dance with
the object of spellbinding the fugitive . 0 The ode is anti-
strophic, with ephymnia, and it is through these refrains,
chanted as they dance round their victim, that the incantation
operates. The ephymnion is used in the same way in the Sup-
pliants, where the daughters of Danaos curse their pursuers
and call down a storm on them while they are making har-
bour . 7 These refrains take us straight back to the mimetic
incantations of primitive magic.
There remains monostrophic form. Here we cannot point to
any tangible relic of the refrain such as we have recognised in
the epode and ephymnion. In this case it has completely dis-
appeared. But it was certainly there once. The proof lies in the
internal structure of the strophe itself, to which I will now appeal.
In the preceding chapter it was argued on general grounds
that the stanza or strophe is universally constructed on the
musical principles derived from the improvisation and refrain
of collective labour. I will now proceed to substantiate this
proposition, in regard to Greek, by a detailed analysis of the
strophe. I shall examine in turn the three oldest specimens that
have survived. Alkman will be represented by his parthcneion or
‘maidens’ song’, Alkaios and Sappho by the stanzas that bear
their names. The discussion will necessarily be somewhat
technical, but I will make it as simple as I can.
The parthitieion of Alkman is a long chorale composed for a
chorus of girl dancers. It is monostrophic, and the strophe is
constructed as follows:
A ton tis otSSv tIotjs* 6 B f fiJkpioj oonj tflippcov
_ u~irir u-uinru
&p{pav SiccttMkei fixXauTOS. ty&v S’&tBco
-uHj-ir uuinru
’AyiSQs t& <j)Cr Apia ‘f wt’ SXiov, Svrrep Spw
iru-ir u-uiru-u
*Ayi6d> uapniprren 9 a(vr|v. — tpl 6* ofrr’tironviiv
“iruir uuinru
* A. E. 307-99. 7 A.Su. 1 18-81.
XV
THE RITUAL ORIGINS OF GREBK EPIC 469
B oCrt pwyifoOcrf vw & kXtwi yppaybs
“iru “iru -iru
oifB’ dpSSj {§• SokeI yip flptv cr>ri
— - —
~U~U “ITU ~U~U
frorpnrfi; toss uomp at tjj Iv PotoTj o-ricretEV tmrow
“iru iru “u“u iru
ircryiv dtOAo^ipov xavoyirroSa. — -r£$v vTr o TrrrpiSfcov ivetpav.
uumruunu “uu-uuiru
For the sake of readers who cannot follow the Greek I give an
English translation in a simple musical setting designed merely
to bring out the rhythmical structure:
ihim* Wing line » fly ing fil- Ip. hght'foot-ed, itur-dy, a ped- i-gr« prizc-wm-ntr.
The strophe falls into two sentences (AB). The first sentence
(A) contains four identical phrases. Each phrase is composed
of two figures, announcement and responsion, one in triple
measure (irCrir), the other in mixed triple and quadruple
47° STUDIES IN ANCIENT GREEK SOCIETY XV
measure (u-uinru). The second sentence (B) also contains
four phrases. The first three are in triple measure, based on the
figure “uni. The fourth begins in quadruple measure (-mruu
-uiruu) and ends with a figure in which triple and quadruple
are again mixed (“uiruirinu), recalling the second figures of
the first sentence (fruiriru). The binary structure of this com-
position is quite plain.
With one exception, all the figures used in this strophe recur
in the other fragments of Alkman. They belong to the common
stock of the convention in which he was working. The excep-
tion is the last figure (-uu-uu-u-ti). This is conditioned by its
context. Its function is to carry the quadruple movement of the
preceding figure into a reiteration of the close of the first
sentence. It is, in other words, a cadence, serving exactly the
same structural purpose as the rhyme in the ballad quatrain.
What is a cadence? ‘That strain again! It had a dying fall’. The
cadence is a lingering echo of the lost refrain.
Let us now try to imagine what would have happened to this
structure of Alkman’s if it had been sung by a soloist without a
dance accompaniment. In these conditions its compass would
have been disproportionate to its function, which is now
reduced to accompanying the voice alone. It would therefore
have contracted into something like this:
A lan ti; aiwv Trials' & 8’ BXpios ootis tGqipoov . . .
-ufiru- u~uu“u"u
’AyiBto tiap-nipETcn 9 a(vtiv. — £p& S’ oOt’ hraivfjv . . .
“truir u-uiru-u
B £icirperrf|S t&s to amp al ns tv PotoTs aT&aeiev tuirov . . .
“ITU nj-U -ITU "ITU
TMV OTTOTTfrpiBlMV ivctpcov.
-utruu-iru
We are reminded at once of the Alcaics we read at school:
A douwJrnui tCv dvlpcov ot&oiv*
U~LTU ~UU“U~
t 6 piv yip JvSev Kupa KuXIvBrron,
uiru
~UU"U“
- XV
THE RITUAL ORIGINS OF GREEK EPIC 471
B t6 C’EvSev fiuutS 8* 6v t 6 iilacov
utj-u nru
vSt <popf|t«ecc crOv ueXaIv$.
“UUTJLTU-U
The storm is raging, scattering all my wits,
The waves are sweeping past to the right and left.
Our anchor lost, our sail in tatters,
Helmless and helpless we drift to shipwreck.
The phrasing of the first sentence is slightly different, but the
strophe falls into two sentences, as before. In the two-figure
phrases of the first sentence triple measure is followed by
mixed, as before. In the second sentence the triple measure is
developed independently, as before. And the conclusion is the
same. I have already remarked that this figure does not occur
elsewhere in Alkman. Nor does it occur in Sappho, nor in
Alkaios except in its present context as the conclusion of the
Alcaic stanza.
These resemblances are too close to be fortuitous. They
show that the Alcaic stanza is derived, not of course directly
from Alkman’s partltncion — that is impossible — but from a
common Graeco-Anatolian prototype. Alkman has preserved
this structure in its older, ampler form, because, as he uses it,
it retains its original choral function. In the Alcaic monody, in
which the dance has disappeared and the chorus has shrunk to a
single individual, the structure has shrunk too, leaving a
masterpiece in miniature.
The Sapphic stanza is of the same small dimensions as the
Alcaic, and its phrasing is similar, but there is one important
difference.
uoiKiXiSpov' dSivar’ ’AippoSha,
ttccT A(os BoXBttAoke, Matronal as,
|t/| n’ fiacticri imjB’ iv[ai<7i B4nva,ir6Tvia, 0O[iov.
Aphrodite, goddess enthroned in splendour,
Child of Zeus Almighty, immortal, artful,
I beseech thee, break not my heart, O Queen, with sorrow and anguish!
The effect of this rhythm, which has always seemed to me one
of the loveliest in poetry, is usually obscured by printing the
last five syllables as a separate verse. It is true that in the Latin
HG. 67 . Alkaios and Sappho: Attic vase
form of the Sapphic die final pentasyllable does constitute a
separate verse. That is dear from the fact that it is often isolated
by hiatus. But these Latin Sapphics were poems pure and simple,
recited not sung, and we shall see in a moment that the isola-
tion of the pentasyllable followed from the loss of a charac-
teristic musical device. In the Greek Sapphic we never find a
hiatus at this point. Moreover, while the first two verses always
end with a word, the third, as usually printed, often runs over
into the fourth. This proves that die pentasyllable is part of
the third verse.
This point is certain, but it leaves us with an apparent
anomaly. Alkman's strophe consists of two sentences, each
containing four phrases. The Alcaic too consists of two sen-
tences, each containing two phrases. The Sapphic opens with a
sentence of two phrases, like the Alcaic, but in the second
sentence, if we treat it as continuous, we seem to have only a
single phrase.
XV THE RITUAL ORIGINS OF GREEK EPIC 473
One of the commonest figures in Greek lyric is the phere-
cra tic (~u-uu~u or u~uu-u). It was a favourite cadence. This
point is fully illustrated in my Greek Lyric Metre. Here it will
be enough to give two examples. The first is from Anakreon: 8
yowoOpcrl a' £Aaq>r|P6AE ?6v6r| At6s, Aypteov 6£cnroiv’ "Ap-npi Onpcou.
-u-uu-D
The second is the refrain of the traditional wedding song:®
'Yprf|V & ‘YpEval’ ‘Yp^v, 'Ypf)v & ‘Yptual* &.
U“UirU
With this cadence in our ears we have no difficulty in analysing
the Sapphic:
A ttoiKAiSpov’ ifldvorr’ ’AfpoSh-a
-iru -uiriru
Tral Alos 5oAo7rXoK£, Afacopal at,
“ITU -uutru
B pfl p' fiaaiai pTlS’ dvfcttm Sdpvo, ir6Ti>ia, 60pou.
-xru -uiriru-uiru
We begin with two identical phrases, as in the Alcaic. Each
has two figures, triple and mixed, as in the Alcaic. The figures
themselves are slightly different. The first (iru) has already
been met with in Alkman. The second (tjuttu) is all but
identical with another of his (u-uinru) and with the phere-
cratic (-truiru). The third phrase begins by repeating the first
two and then concludes the whole by passing into the phere-
cratic cadence. Thus, the disyllable Sotuva has a double value.
It completes the repetition and introduces the pherecratic.
This is the device familiar to musicians'under the name of
overlap.
So the second sentence contains two phrases after all. And
now we see that the Sapphic was evolved from the primitive
sequence of solo and refrain by merging the second element,
the refrain, into the first and so investing it with the value of
a cadence. As I said before, the cadence is an echo of the lost
8 Anac. I.
®The refrain had many forms: Ar. Av. 1743, ^ a ‘ 1 33 2 > E. 2V. 314-31,
Theoc. 18. 58.
474 STUDIES IN ANCIENT GREEK SOCIETY XV'
refrain. It would be hard to find a more perfect example of
a ritual form transmuted into art. 10
3. The Hexameter
Our argument has now gone far enough to give us a lead on
the origin of the metre of Greek epic, the dactylic hexameter. 11
The nucleus of poetry was the verbal element in the undif-
ferentiated complex of primitive song and dance. As the kernel
grew, the shell decayed. First the dance was shed, and then the
music. And the rhythmical form was simplified. We have
seen the strophe contract. We shall now see it disappear.
The dactylic hexameter must be studied in conjunction with
the trochaic tetrameter and the iambic trimeter. For the sake
of brevity I shall refer to these three metres as the hexameter,
tetrameter, and trimeter.
The tetrameter and trimeter first appear in the fragments of
Archilochos, whose floruit may be placed in the latter part of
the eighth century. They were both used by Solon. The
tetrameter was adopted by the early dramatists as the medium
of tragic dialogue, but later it was superseded by the trimeter,
which according to Aristotle was nearer to the rhythm of
common speech. 13
The structure of these metres is as follows:
Hexameter: -uiruir/u/u-uu-uiHi
Tetrameter: nru-iru/"iru-uu
Trimeter: u“u _ u/-u/”u-uu
13 1 take this opportunity of drawing attention to what seems to me
a major weakness in my GLM — my failure to discriminate between
different periods and schools of lyric: (1) the Graeco-Anatolian school — Alk-
man, Sappho, Alkaios; (2) the western school — Stesichoros, Ibykos; (3) the
mature convention of Simonides, Bakchylides, Pindar, and the dramatists.
My distinction between 'Dorian', 'Aeolian', and 'Ionian' is fully applicable
only to the last.
11 On this subject I have reached by a different approach much the
same conclusion as Bergk and Usener, whose theories are now generally
abandoned. Bowra’s view, that the source of the hexameter 'must be a
primitive type of narrative poetry whose unit was not the stanza but the
line’ (TDI 61-2), simply shelves the problem.
12 Arist. Poet. 4. 18-9, cf. Demetr. 43.
XV THE RITUAL ORIGINS OF GREEK EPIC 475
They are often treated simply as sequences of so many dactyls
(~uu), trochees ( - u) or iambs (ir). This analysis indicates the
time and the length of the verse, but it tells us nothing about
its organic structure. That depends on the internal break in the
words, or cssura, which in the hexameter and trimeter always
falls in the middle of a foot. The foot is an abstraction, with no
organic value, like an isolated step in dancing or the isochronous
bar in music. The organic unit is the figure, representing a
series of steps or beats, which functions as a single unity and
not as an aggregate of its parts.
Before we proceed, I must explain the tests we apply to a
piece -of Greek lyric in order to distinguish the figures and
phrases of which it is composed. 13 There are three: the division
of words, hiatus, and the use of irrational syllables. If we find
a break in the words at the same point in the strophe each time
it is repeated; if a word ending in a vowel is followed by one
beginning with a vowel, with each vowel functioning in-
dependently in the metre; if a long syllable is substituted for a
short or a short for a long: all these signs normally occur at the
junction of two figures, which is of course in origin a musical
rest, corresponding to a pause in the dance. Examples will be
found in the strophe from Alkman quoted above.
With this in mind, we are able to see that the opening figure
in that strophe is really a compound figure. This is shown by
the quantity of the fourth syllable. In the first, second and
third verses this syllable is short ("uvir) but in the fourth it
is long (~u — u _ ). In other words, this figure is composed of
two originally independent elements “u - u and "u~. The first
is used separately in the second sentence of the same strophe
(oOte uwnqcrQal viv & .Khevva x°pay6s) and in the Alcaic
(ov t6 p&j-ctov) and the Sapphic ( 7 roiKiA 60 pov’). The second
is common in the work of Alkman and other early poets:
Alcm. Zi K&rrwp t* ttcWvo:v d*£wv.
Alcm. 6i: oOBJ tSI KvccKiAu oiiSI rd NuptrtXot.
Returning to our three metres, the first characteristic that
distinguishes them from lyric is that they are monophrastic,
13 The MS divisions date only from the Alexandrian period: in earlier
times lyric poetry had been written continuously, like prose.
476 STUDIES IN ANCIENT GREEK SOCIETY XV
that is, they consist of a single verse continuously repeated.
Secondly, they are isochronous. The hexameter is in quadruple
time, the other two in triple time. There is no mixed time.
Their structural uniformity was conditioned by the mode of
delivery. The poems composed in these metres were recited.
The whole attention was concentrated on the spoken word.
That is why the metrical pattern is so simple. But this sim-
plicity was not crudity. On the contrary, after being evened out
in this way, the pattern was used as a keyboard for rhythmical
subtleties of a new kind, which were precluded in the strophe
by its structural diversity. It became possible to elaborate an
endless variety of verse paragraphs conforming to the natural
flow of common speech.
One of the salient features of Homeric Greek is its wealth
of polysyllables. In later Greek, especially Attic, these were
reduced by the contraction of contiguous vowels. The change
had a marked effect on the rhythm of the language, as may
be seen by translating a piece of Homer into Attic:
6 cripiv tii qipoviuv dyop^crerro tcotl tiET&nrtv.
6 aipiw tO ippovcov fiyop^aaro Kal meteIttev.
The Homeric verse is dactylic; in Attic it becomes trochaic.
This must have been a factor in the decline of the hexameter.
As the language became less dactylic, this metre lost its
vitality and gave place to others, closer to speech.
In the hexameter and trimeter, the caesura, which always falls
in the middle of a foot, has two alternative positions: in the
former, before or after the second syllable of the third dactyl;
in the latter, in the third or fourth foot. It is the movable
caesura that makes these metres so flexible. Not only is every
verse broken up into two units conflicting with the time
pattern of the whole, but each succeeding verse, though
metrically identical with the last, can be made rhythmically
different. This perpetual interplay of like and unlike is the
life and soul of the metre.
The tetrameter lacks these advantages. The caesura has only
one place, and that is at the end of a foot. Since it always coin-'
cides with the time pattern, it makes the rhythm less flexible
and so more obtrusive. It is, as Aristotle says, too 'dance-like'.
XV THE RITUAL ORIGINS OF GREEK EPIC 477
The dramatists abandoned it for that reason. They preferred
the trimeter, which, with its triple time and its movable
cicsura, was the nearest of the three to common speech.
What was the origin of the c.-csura? We see that it is vital
to the rhythmical effect, yet there is no reason to suppose
that it was created in response to any inherent natural need.
There arc many metres, in Greek and other languages, that
have no ciesura. Its origin must be historical.
The tetrameter, being the crudest of the three, is likely to
be the most self-revealing in its structure.
Ci (laOirjcdVcov fivcraacc TTcpalSuv CnTtpTirrT),
-iHHru “unHr
liiyrcp f| 3fp£ov ytpaid, xaTpc, Aapdov yirvai.
-iririru -u-uir
Obscrving the incidence of irrational syllables, we get the
following formula nnj-irO /-tru-ir. This is a phrase of two
figures, which differ from one another only in the ending,
while the second is identical with the opening figure in Alk-
man’s strophe (TTepolScov CnrepT<5rrT| — Ion tis cncov t(ctis). The
caesura was derived from the internal break in a two-figure
phrase.
The trimeter has irrational syllables in the first and last
places, also in the fifth and ninth. This gives us another two-
figure phrase, divided at die commoner caesura, urrv/Trurr:
0£ous uh> alT&S tQvS’ &rraMayf|v irtvcisv
(jpoupas htlas IitJkoj, f|V Kom&ptvos. . .
The second of these figures is the one we have just identified in
the tetrameter (toovS* d-rrc&Aocy f) v tt6vcov— TT epolScov vnrepT&TT)—
fan ns aiuv Tims). The first is the opening of the Alcaic
(QeoOs pfcv alTcS — <5towvhT|pi).
The second figure, as we have seen, is a compound one
(-VTJ/-U-). The trimeter is thus constructed of three elements.
All three are common in lyric, separately and in similar com-
binations:
Alcae. 65: St|cn pe KcopdgavTa, Stfai, Macropal Of, Mooopcn.
Alctn. 2: Kdorcop te irtiXcov Catdav Spcrriipes IrnnSrai ooipo!.
Anac. 79: {pto te 6t)0ts koOk iptu Kal itafvopcn koO palvopcn.
478 STUDIES IN ANCIENT GREEK SOCIETY XV
Examples might easily be multiplied, but there is no need to
labour the point. Seeing that the combinations u-u-u/ir and
-irU/nj- are common in lyric from the earliest times, we
cannot doubt that the trimeter, composed of the same elements,
was a combination of the same kind.
One of the features of epic diction, which will be examined
later, is the use of stock phrases. Many of these are very old.
They are the stuff of which epic poetry was made. Most
of them consist of a half-verse preceding or following the
caesura. One of the commonest types of Homeric verse is con-
structed of two such phrases, divided usually at the feminine
caesura:
t6v S' ebj o Ov fv<Sr]C7EV dpntylXoj Mevftaos.
t6v 5* cxOte uporfemE -rroXiWiccs BTos ’OSwa&is.
c5>S ?9orr' ou5* &irt0ti<re fled AeukcoXevos "Hpti.
The formality of these set verses, used repeatedly and without
variation wherever the subject requires them, stamps them as
archaic, and suggests that the hexameter too arose from the
combination of two figures, the break between them surviving
in the caesura. What the original figures were is a more difficult
question. The hexameter is the oldest of Greek metres and its
early history is lost. On the main point, however, we may
claim with some confidence that it originated in a two-figure
phrase of a type which is known to have dominated early
Greek lyric and is based on the announcement and responsion
of binary form.
This conclusion is confirmed by evidence of a different kind,
which makes it certain that, whatever the intermediate stages
may have been, the verse-form of Greek epic is descended from
choral lyric.
In the historical period epic poetry was unaccompanied.
The minstrel declaimed it, holding a staff in his hand. Hesiod
alludes to the minstrel’s staff, and there is a story that he was
once defeated in a minstrelsy competition because he was
unable to accompany himself on the lyre. 14 In the Iliad and
Odyssey there are several descriptions of epic recitals, purporting
to refer to the heroic age. In these the minstrel invariably
14 Paus. xo. 7. 3. Cf. Murko 285.
XV
THE RITUAL ORIGINS OF GREEK EPIC 479
sings the lay and plays as he sings on the lyre. The staff of later
times was evidently a ritual substitute for the lyre.
We had no difficulty in tracing the trimeter and tetrameter
back to the art of song, but we did not establish any direct
connection with the dance. In epic it is clear.
When Telemachos arrives at Sparta in the fourth book of the
Odyssey, Menelaos is celebrating a wedding. A minstrel is sing-
ing to his lyre, and meanwhile two trick-dancers are spinning
in and out of the crowd, singing as they go. 1 ® In Phseacia we a
witness an entertainment of the same kind:
The herald went to fetch the minstrel's lyre from the palace. Meanwhile
nine judges, appointed by the people to administer the competitions, got
up and prepared a dancing ground, smooth and wide and level. Then die
herald arrived with the lyre and handed it to Demodokos. He sat in the
middle of the ground, while a group of young men, trained dancers in the
first flower of manhood, took up their positions round him and began to
dance divinely. Odysseus gazed in astonishment at their flashing feet, while
the minstrel struck up with his lyre and sang of the amours of Ares and
Aphrodite. 1 ®
Here we have the art of epic in its original setting.
The evolution of Greek poetry out of primitive ritual, in
which song and dance were combined, has now been demon-
strated by a concrete analysis of its metrical forms, and the
result confirms the conclusions reached in Chapter XIV regarding
the origin of poetry in general. Having established the history
of the performance, we turn our attention to the performers.
It can be shown that the epic poet stands in the same relation to
the singing and dancing chorus as the epic hexameter to the
complex of song and dance.
4. The Chorus
Greek state religion was founded on the clan cults of the big
landed families which set themselves up in the p 6 lis (p. 3 5 ^)*
Each family fostered its own cult as a means of enhancing its
prestige, while their collective monopoly of worship secured
their position as the governing class. After the democratic
revolution the cults were brought under state control, though
in many cases their administration was left in the hands of
i® Ol 4. 17-9. 10 01 8. 256-67.
480 STUDIES IN ANCIENT GREEK SOCIETY XV
their hereditary owners. Thus, with very few exceptions, the
clan cults are known to us only to the extent that they became
state cults, and of course in that process they were transformed.
The clan cults that survived as such, being private, have left
little trace in our records.
In spite of this we can discern at least the outline of the old
clan ritual. It was the archetype of the choral ode. Choral
lyric was as characteristic of the aristocracy as drama was of
the democracy and epic of the heroic monarchy. Its technique
was a heritage from tribal times, handed down through these
conservative families with relatively little modification. Hence,
although it only came to fruition after epic had passed its
prime, it was structurally more archaic.
Its aristocratic character is apparent in even its latest mani-
festations. Pindar lived at a time when, except in Sparta, Elis,
and Thessaly, the old nobility had been forced almost every-
where to come to terms with democracy. All his extant odes are
composed for prizewinners at the athletic festivals. These
gatherings attracted crowds of hucksters and holiday-makers,
but the games themselves were aristocratic. Only the well-to-do
had time for gymnastic training, and the most coveted of the
prizes, for the chariot race, was in effect reserved for landed
gentry with a tradition of horsemanship.
The ode was designed as an ovation for the victor on his
return to his home town. It was composed by a professional
poet, who took no further part in the proceedings, but its per-
formers were a chorus of the victor’s kinsmen, accompanied
by an instrumentalist. It was an encomium, a song 'of praise
for the prizewinner, and of course for his family as well. In the
typical Pindaric ode the praise of the individual comes at the
beginning and the end. The centre is reserved for a myth,
which is taken in many cases from the actual traditions of the
victor’s family or clan. Such was the procedure in Pindar’s
day, but the employment of a professional poet was an innova-
tion. For earlier times we must envisage a chorale composed
as well as performed by the victor’s kinsmen — a hymn of
praise to the clan.
Pindar composed many other types of ode, all now lost —
hymns, processionals, paeans, dithyrambs, dirges, parthineia.
XV THB RITUAL ORIGINS OF GREEK EPIC 481
His dirges in particular would have been interesting. In Greece,
as elsewhere, this ceremony was immemorially old, and like the
athletic ode it was cultivated by the nobility for the sake of
family prestige. In several states we hear of sumptuaiy laws
limiting the size, duration, and expense of funerals. 1 ’ At
Athens they dated from Solon. Their object
was not simply to discourage private extrava-
gance. They were directed against the clans.
When a man had been killed in a brawl, his
whole clan followed his body to the clan
cemetery, where they were worked up into a
frenzy by the dirge performed at the grave-
side. The result was a vendetta. The explosive
character of these occasions finds an echo in
the Oresteia, where Agamemnon's children,
standing at his tomb, pass from singing his
praises to a furious clamour for revenge . 18
In its primitive, pre-artistic form the dirge dancer: gem from
was performed by women. That is why, in these - Vaphclo
sumptuary laws, only a specified number of women, within
certain degrees of affinity, are permitted to enter the house of
the dead, and various restrictions are placed on their behaviour
at the graveside. The women, clinging tenaciously to the customs
of the past, were the worst mischief-makers. Their inferior
status in patriarchal society left them with a traditional disrespect
for law and order.
The structure of the Pindaric dirge is unknown, but we have
an earlier example in the lament for Hector at the end of the
Iliad. He is addressed in turn by Andromache, Hecuba, and
Helen, and after each has spoken cries are raised by the other
women in attendance. The three speakers are the leaders, the
others form the chorus . 19 The ritual features of the performance
« Plu. Sol. 12, D. 43. 6a, S 1 G. 1118-9, GDI. 1561.
18 A.C. 305-476.
19 Ii. 24. 7x9-76. There is a discrepancy in this passage. The three
women are described as ‘leading the dirge' (723, 747, 761) the other
women supplying the refrain (74Q; but we have been told at the beginning
that the leaders are male minstrels (720-2). I take it that the primitive
female dirge has been confused with the later professional type, cf. Chad-
wick GL 3. 61.
Gg
482 STUDIES IN ANCIENT GREEK SOCIETY XV
are characteristically slurred over, but we know from other
sources that it was commonly sung, not spoken, and accom-
panied with an ecstatic dance in which the women beat their
breasts and tore their hair . 20 This is the world-wide form
of the dirge at the present day . 21
The loss of Pindar’s parthineia is compensated to some extent
by the survival of one of Alkman’s. This is the piece whose
musical structure we examined in the last section. It contains
some valuable information.
The occasion was probably the presentation of a new robe
for the image of Artemis . 22 Annual investitures of this kind
were common. In the Iliad we read of a brilliantly embroidered
robe conveyed by a procession of women to the Trojan Athena . 23
At Athens, on the 21st of Thargelion (May-June), the image
of Athena Polias was veiled, taken down, and washed . 24 It was
then clothed in a new robe woven by the arrhephoroi (p. 222 ). 26
The proceedings were in charge of the clan Praxiergidai . 28 The
day on which the image was dismantled was a dies nefasta, one
of the blackest in the Attic calendar. 2 ’ Some authorities have
supposed that the day got its bad name from the ceremony , 23
but this cannot be, because the same day is a bad one in
Hesiod’s Works and Days . 29 It was bad because it fell properly
in the period of the waning moon, and the ceremony associated
with it was originally a monthly rite of purification.
Among the priestesses of Artemis at Ephesos was one called
the kosme'teira. Her office was hereditary, and her title implies
that she was in charge of the investiture . 30 When this took
place we do not know, but it was probably another rite of
purification or regeneration. In the cult of the same goddess at
28 II. 18. 50-1, A.C.423-7, Per. 123-8, 1039-77, 5k. 126-8, cf. Th. 2. 34.4.
21 Bucher 442.
as I suspect that the robe was embroidered with stars, cf. II. 6. 295,
Orph. fir. 238. This would make 60-3 intelligible: 'The rising Pleiads vie
with us as we bring to Orthia a robe like Sirius through the fragrant night',
i.e. it outshines the Pleiads.
23 If. 6 . 286-303. 24 piu. Alcih. 34.
2® Harp. s.v. AppqipopsTv, Ar. Av. 826-7. 28 Plu. Alcih. 34, Hsch. TTpagicpylSai.
27 Plu. I.e., X. Hell. j. 4. 12, Poll. 8. 141. 28 Deubner 22.
28 Hes. Op. 803. so SIC. 1228, CIG. 2823.
XV THE RITUAL ORIGINS OF GREEK EPIC 483
Brauron the robe was made from the clothes of women in
childbed.* 1
Alkman’s parthltuion was performed by a chorus of ten or
eleven unmarried girls. For them, as for their goddess, it was a
rite of regeneration or initiation. Probably they formed an
agila. This was a sodality of candidates for initiation. It is
known that girls as well as boys were organised in this way. 32
The members of the male agila were all of common descent in
the male line (p. 145); and in this ode die two leading girls are
described as cousins. It seems clear therefore diat the agila was a
kinship group within the clan, and, if so, the rite in which
these young ladies arc engaged is a clan cult.
The accompanist was Alkman himself. This appears from
other fragments of his work in which he takes part in the
singing — as for instance where he sings a playful apology to
his chorus girls for being too old to join in die dancing. 33
This parthlneion is die oldest choral ode we possess, but there
is reason to believe that odes of this type had been cultivated
for centuries before Alkman. The Greeks themselves recognised
that there had been ‘poets before Homer', and they mentioned
names. Two at least of these must have some historical foundation.
The name of Pamphos, an Athenian, was preserved by a
hereditary cult society of women, die Pamphidcs. 34 His work,
which is said to have influenced Sappho, included hymns to
Demctcr, Persephone, and die Charitcs. 30
Olcn was a Lycian who setded in Delos, where he composed
several hymns to Apollo. 30 One of these was danced by a
double chorus of boys and girls who had just reached puberty. 3 ’
Anodier was addressed to the goddess of childbirth. When
Lcto was in labour, seven songs were ‘sung over her’ by swans
from Anatolia, while nymphs of Delos uttered the 'sacred
chant of Eilcidiyia’. 38 Olen was also credited with a hymn to
Hera and with die invention of the dactylic hexameter. 3 0
31 E. IT. 1450-67. For other similar investitures sec Hyp. 4. 25, Paus.
3. 16. 2, 3. 19. 2, 5. 16, 7. 23. 5, IG. 5. 2. 265. 19.
32 Pi. fr. 112. 33 Alcm. 94. 34 Hscli. notu<pI 5 ey.
30 Paus. :. 38. 3, :. 39. 1, 7. 21. 9, 8. 35. 8, 8. 37. 9, 9. 27. 2, 9. 29. 8,
9.31.9,9.35.4-
30 Hdt.4. 35, Paus. 8. 2i. 3, 9. 27. 2, Suid. 'toViv.
33 lb. 249-57. 30 Paus. 2. 13. 3, 10. 5. 7.
3 ’ Call. HDel. 296-9.
484 STUDIES IN ANCIENT GREEK SOCIETY XV
The cults to which these hymns refer were all of matriarchal
origin, and the hymns themselves ' seem to have been per-
formed, like Alkman’s, by a female chorus under a -male
leader. This point is confirmed by one of the most familiar of
all Greek traditions. The arts of music, dancing, and poetry
were under the patronage of Apollo and the
Muses. Pindar tells how at the marriage of
Peleus and Thetis the Muses danced in
a chorus led by Apollo with his seven-
stringed lyre . 40 Alkman invokes the Muses
to sing for his girl dancers a new song . 41
An ode attributed to Terpandros begins
with an appeal to the son of Leto, leader
of the Muses . 42 In the Iliad, when the
Olympians have fed and the wine is going
round, Apollo plays his lyre while the
Muses sing ‘answering one another', that
is, antiphonally . 40 In the Homeric Hymns they
sing, again ‘answering one another’, of the
deathless gifts of the gods and the sufferings of
mankind, while the Horai and Charites dance
hand in hand to the music of Apollo, who takes
a turn himself in the dance . 44 This lyre-playing
god with his chorus of goddesses is simply a
celestial reflex of the prehistoric partheneion.
. „ On one occasion he is absent — at the funeral
FIG. 69. Apollo r aImi
and lyre: Attic 0t Acnmes -
vase The Nereids stood round the body, weeping bitterly
as they wrapt it in the winding-sheet, while the nine Muses, answering
one another, sang the dirge. 40
For the Greeks it was enough to say that the god of light and
health could not have been present here because he was by
nature incapable of lamentation. This was proverbial . 40 But
in reality his incapacity was an effect and not a cause. He could
40 Pi. N. 5. 22-5, cf. Hcs. Sc. 201-6, Paus. 5. 1 8. 4. 41 Alcm. 7, cf. 68, 94.
42 Tcrp. 3 Bergk. 40 II. j. 603-4.
44 Horn. H. 3. 188-201. « Od. 24. 60-1, cf. II. 18. 50-1.
40 A.A. 1058-63, Per. 608, E.Sk. 971-9, Stes. 22, Sa. 109.
XV
THE RITUAL ORIGINS OF GREEK EPIC' 485
not mourn because he stood for the male chorus-leader, and
the prehistoric dirge had been performed exclusively by
women.
In respect of divine patronage no distinction was drawn
between choral lyric and epic. All the extant epics begin by
invoking the Muses. When Odysseus wishes to compliment
Demodokos, the minstrel of Phseacia, he says he must be in-
spired by Apollo or the Muses. 47 Further, at this level the poet
merges into the priest. The minstrel's person was sacred. Being
inspired by Apollo, possessed, he was a prophet as well as a
poet. He claimed to. know the future. 48 So with Apollo himself.
He combined prophecy with music because in primitive-
society music is the vehicle for all forms of psychical dissocia-
tion, including the prophetic trance. He was at once poet,
prophet, priest — the male priest of a female cult.
How old was this concept of Apollo and the Muses? It
was certainly prehistoric; yet, if we press our analysis further,
the divine chorus falls apart. The Muses came from the
north — from Mount Helikon in Boeotia and from Pieria on
the slopes of Olympus. 4 ® Their name probably means ‘mad
women’. 60 They resolve themselves into a female thiasos of the
same type as the Bacchants, the ecstatic votaries of Dionysus,
who also came from the north. 61 In historical times their
main centre was Thespiai, where they were worshipped by a
society named after Hesiod. 68 They were not prominent at
Delphi, and at Delos their place was taken by the Deliades
and Minoides. 68 Apollo came from the south — from Crete
(pp. 293-4), where his distinctive instrument, the lyre of
47 Oi. 8. 487-8, cf. Horn. H. 25. 2-3.
48 Od. 8. 479^8 x, 22. 345-6, Hes. Th. 31-2, II. 1. 70.
4 ® Hes. Th. 52-3, Str. 4x0, 471.
60 Roscher LGRM 2. 3238.
61 The Mousai actually figure as votaries of Dionysus in the Agriania
at Orchomenos: PIu. M. 717a, cf. Erat. Cat. 24. 140.
68 Paus. 9. 31. 4, IGScpt. 1785, 4240, cf. 1735, 1760, 1763. The
name Thespiai is properly a cult title of the Mousai (see p. 129), and it is
possible that here and elsewhere in Boeotia they had replaced the Charites,
whose worship was very ancient at Orchomenos (Paus. 9. 35. 1; Roscher
LGRM 1. 877-8) and probably of Syrian origin: Gaster GSF.
68 Evans PM 3. 74.
486 STUDIES IN ANCIENT' GREEK SOCIETY XV
seven strings, can still be seen on the sarcophagus from Hagia
Triada . 84
For these reasons it may be suggested that this concept
crystallised on the Greek mainland, perhaps in Boeotia, under
Minoan influence and in response to a definite stage in the-
decline of the matriarchate, marking the point at which cults
previously reserved to women were brought under the control
of a male priest.
If the leader of this female chorus was an intruder, how did
he gain admittance? By disguising himself as a woman. That is
what Pentheus did when he went to spy on the Bacchants . 66 It
was a common thing at Dionysiac festivals for the men to dress
in women s clothes . 68 The costume of the Lydian priesthood
was properly a woman's. 8 ? This should not surprise us. Rather,
we should look with a critical eye at the mitres, stoles, and
frocks of our own clergy. All over the world the transfer of
religious authority from the one sex to the other has been
effected by dressing the priest as a priestess . 68 The motive was
- partly no doubt to make the change acceptable by pretending
there had been no change at all, but there was more in it than
that. The traditional costume was sacred, charged with magic,
and therefore indispensable.
One of the frescoes at Knossos depicts a festival in an olive
grove . 69 In the right foreground a chorus of fourteen women
are dancing. They are moving towards the left with extended
arms. Behind them is the audience. Immediately behind, the
dancing ground there are groups of women sitting on the grass
and chatting. Behind these, separated by a barrier, is a crowd
of men, closely packed, all standing, all intent on the per-
formance. It seems that the men are mere spectators, while the
women on the grass are participants who later on perhaps will
take their turn at the dance. The left foreground, to which the
dancers are pointing, is missing, but there. is not much doubt
what it contained. On a gold signet ring of the same period we
see three women dancing in a field of lilies . 60 Two of them
have their arms raised; a fourth, standing on a higher level, has
64 Evans PM 2. 834-6: fig. 73. 65 E.B a. 821-36. 66 Luc. it cal 16.
67 Ramsay 174. 68 Briffault 2. 531-6. 69 Evans PM 3. 67-8.
00 Ik 3. 68: fig. 71.
XV
THE RITUAL ORIGINS OF GREEK EPIC 487
one band on her hip and die other at her forehead (p. 243).
Still higher is a fifth figure, also female, but dressed in an
archaic costume and separated from the others by a broken
wavy line. This, as Evans has shown, signifies the boundary
between earth and heaven. In response to her votaries, a female
HG. 71. Descent of tie goddess: Mirtoan signet
chorus with a female leader, the goddess descends to inspire
them with the ecstasy of the dance.
Both fresco and ring arc assigned by Evans to Middle
488 STUDIES IN ANCIENT GREEK SOCIETY XV
Minoan III. The type persisted. We meet it again in a Late
Minoan terracotta model. Three women with outstretched
arms are dancing in a ring, while a fourth sits in the middle
and plays a lyre . 61 The Minoan chorus seems to have remained
completely feminine — 'a symptom', as Evans remarks, 'of the
fig. 72. Minoan chorus: terracotta
matriarchal stage '. 62 Only in Late Minoan III can we detect
signs of a change. The procession on the Hagia Triada sarco-
phagus includes a young male lyre-player — a budding Apollo.
We know his sex by the colouring of his skin . 68 Otherwise we
should certainly have taken him for a woman, because he is
dressed in a long robe reaching to the ankles with a bodice open
at the breast, exactly like the girl in front of him. He is
dressed as a woman because he is performing a woman’s task.
Starting with Alkman, we have argued back from the female
chorus led by a male leader to a female chorus led by a female
leader — from poet to priest and from priest to priestess. What
was the minstrel’s place in this development?
In the historical period the minstrel, or rhapsode as he was
61 Evans PM 3. 73: fig. 72. os ft, 3. 75. g3 Jh. 2. 836.
FIG. 73 . Minoan lyre-player: Hagia Triada Sarcophagus
49 ° STUDIES IN ANCIENT GREEK SOCIETY XV
then called, was merely a professional reciter, not a musician or
a creative poet. But in earlier times he had composed his own
lays and provided his own accompaniment. This is the stage
represented by the traditional figure of Homer, ‘the blind
bard of rugged Chios’. Homer sang of the-past, but the min-
strels he describes in his poems sing impromptu of con-
temporary events. And on one occasion, as we have already
FIG. 74. Mixed chorus: Attic vase
noted, the subject is non-heroic, a lay of Ares and Aphrodite,
sung as an accompaniment to a dancing chorus. Thus, as we go
back into the past, epic merges into choral lyric. But there is
one link still missing. In the Iliad and Odyssey the minstrel and
his dancers are always male. Mixed dances are mentioned, but
without a minstrel. 64 In order to meet this point we must take
up the problem from another angle.
5. The Epic Prelude
Prizes for epic recitals were offered at all important festivals.
Professional minstrels travelled from city to city, competing
wherever they went. 8 6 Their great centre was Delos. 6 6 The Delian
festival of Apollo was attended by crowds of both sexes and all
ages, together with choruses entered by different cities for the
musical events. Homer himself is said to have competed there.
The recitals were preceded by prefatory hymns (proolmia ) of
the type that survives in the collection known as the Homeric
Hymns. Most of these can be dated to the seventh and sixth
64 II. 18. 567-72, 590-606. 66 PI. R. 6ood, Io 541b, Certamen 55.
ee See below p. 551.
XV
THE RITUAL ORIGINS OF GREEK EPIC 491
centuries. They vary in length from less than a dozen verses to
several hundred. They are addressed to various deities and
were doubtless intended for different festivals. They are com-
posed in the epic metre and the epic dialect. They differ from
the epic poems proper only in the nature of their subjects. The
epics deal with heroic themes — the 'glories of men, whereas
these preludes are all devoted to legends of the gods. The dis-
tinction was traditional. In the Homeric Hymn to Helios,
after praising the god, the minstrel concludes:
O Lord, farewell! Grant of thy grace a life after my own heart, and now,
having begun with thee, I will sing of the mortal race, the heroes, whose
deeds the Muses have made known to man.®?
Putting this evidence together, we can reconstruct die pro-
cedure. The competition opened with a prelude, a hymn to
the deity of the Occasion. Then the first candidate came for-
ward, took the staff (rhdbdos) in his hand, and began the Iliad .
When he had finished the piece set for him, the next candidate
took the staff and continued.
These competitions for reciters had been preceded by competi-
tions for poets, composing as diey competed.
The tradition of such a contest between Homer
and Hesiod survived in various forms, which,
though apocryphal, are good evidence of pro-
cedure. In one of tire Hesiodic fragments we
read: 'Homer and I were the first minstrels
to sing in Delos of Apollo son of Leto, stitch-
ing our song in new hymns'.® 8 The rivals had
improvised by turns, one chanting, the other
silent, while the lyre or staff, as it passed
between them, marked the ins and outs of
what was in effect a continuous composition.
Hence the term rhapsode (rbapsoidds), which
means properly a ‘song-stitcher’.® 8
In the procedure outlined above there is a
clear-cut division between the sacred and
secular stages, between the prelude and the
epic proper, but there is reason to believe that in early
times they had been continuous. In the Hymn to the Delian
®? Horn H. 31. 17-9. 66 Hcs. fr. 265. ® 8 Cf. Bowra TDI 41.
HG. 75-
A rhpcle:
Attic vase
XV
492. STUDIES IN ANCIENT GREEK SOCIETY '
Apollo the author, doubtless one of the Homeridai, presents
himself as Homer, and describes the festival as it had been in
Homer's day:
How marvellous they are, how memorable, the girls of Delos, Apollo’s
handmaids, who after uttering a hymn of praise to Apollo, Leto, and
Artemis, recall in song the men and women of old. 7 ®
This seems to point to a time when the second stage had fol-
lowed the first without a break. And here, somewhat unex-
pectedly, we are brought back once again to Alkman. His
parthfneion is not complete, but nearly so. It begins with a
myth — the fight between Herakles and die sons of Hippokoon; it
goes on to the task in hand, the presentation of the robe, and
the second part is taken up with a gay exchange of repartee
between the dancers. We recognise the same structure — die
divine exordium, the human sequel.
This sequence lies very deep in Greek poetry. It was a
proverbial rule that the poet 'began with God’. 'O Hymns
diat sway the lyre’, says Pindar, ‘what god, what hero, what
mortal shall we celebrate ?’ 71 Describing the Muses’ wedding
chant for Thetis, the same poet says that ‘they began with
Zeus and then sang of Thetis and Peleus ’. 72 Alkman has the
same formula: ‘I will begin with Zeus and sing’, and Ter-
pandros had used it before him: 'O Zeus, beginning of all,
leader of all, to thee I dedicate the beginning of this hymn *. 78
The earliest examples of all — from the Iliad and Odyssey — will
be mentioned in a moment. The Pindaric ode is an apparent
exception. It begins and ends with the victor. But the poet
usually contrives to combine this scheme with an introductory
appeal to. the gods . 74
We can now see how the procedure at Delos had grown up.
The epic recitals were intrusive. They were established there
during the expansion of the Homeridai. Previously there had
been simply a choral hymn devoted successively to gods and
mortals. The Homeric recitals did not altogether supplant the
hymn, but they absorbed its secular portion, and so it shrank
into a preface, a purely formal inauguration of the real business.
78 Horn. H. 3. 156-61. 71 Pi. O. 2. 1-2, 72 Pi. N. 5. 25-6.
73 Aicm. 9, Terp. 1, cf. Xenoph. 1. 13.
74 Pi. O. 2. 1-5, 3. 1-4, 4. 1 -10, 5. 1-3 etc.
XV THE RITUAL ORIGINS OF GREEK EPIC 493
Tu rning to the Iliad and Odyssey, we find vestiges of the same
sequence. Both begin with an appeal to the Muses. It is very
brief— a mere formula for introducing the narrative. 78 But the
fact that it is there at all raises a suspicion that the heroic lay
itself had once begun with something in the nature of a hymn.
This is confirmed by Hesiod.
The appeal prefaced to the Works and Days has an independent
unity of its own — a miniature hymn to Zeus. And in the
Theogony the exordium runs to more than a hundred lines,
longer than most of the Homeric prooimia. In it we are told that
the Muses sing first of the race of gods, then of Zeus, father of
gods and men, and finally of the Giants and mankind. The
body of the poem expounds the origin and history of the gods,
but this is followed by an enumeration of the goddesses who
mated with mortals, and the poem ends with these words:
‘And now, Olympian Muses, daughters of Zeus, sing of the
race of women*. The sequel, the so-called Catalogue of Women,
survives only in fragments, but the conclusion of the Theogony
shows that the two poems were designed to be taken together,
expounding in turn the history of the gods and the history of
human heroines. Thus the theme of the Catalogue is a heroic
one, but concerned with heroines instead of heroes. Its material
is pre-Homeric, dating from the prehistoric matriarchate. And
with the Theogony it constitutes a hymn of the same structure as
the one in which the girls of Delos celebrated first Apollo and
Leto and then the men and women of the past. The Hesiodic
school was less secular than the Homeric, and so the old structure
survived.
Thus it appears that in content as well as form choral lyric
and epic rest on a common ritual basis. The same thematic
sequence has been found underlying both. We must now
try to define more closely what this sequence is.
78 The preface to the Oiyssey ends (1. 10): 'Take up the story from
thereabouts and tell it to us’. The suggestion intended is that it has been told
by other minstrels many times before (hence the emphasis) and that the
present rendering opens more or less at random, cf. 8. 500, 1. 492, 8. 493.
The invocations prefaced to the Catalogue of Ships and the Exploits of
Agamemnon (II. 2. 484-92, 11. 218-20) are signs that these had been
current as separate lays.
494 -STUDIES IN ANCIENT GREEK SOCIETY- XV
6. Songs after Supper
The minstrel performances described in the Odyssey fall into
two classes, choral and non-choral. The choral examples were
performed on special occasions — in the one case a wedding, in
the other an entertainment for Odysseus. On the non-choral
occasions the procedure was more uniform.
There are three of these non-choral lays, two by Demodokos
and one by Phemios. The former sings of the quarrel between
Odysseus and Achilles, and the Wooden Horse; 76 the latter of
the return of the Achaeans after the war. 77 All these themes are
heroic, and they are all sung after the evening meal. This was
the rule:
What a fine thing it is to hear such a minstrel, with a voice like a god’s! I
know of no occupation more delightful than this — when the tables are
loaded with good things and the wine just poured out, to sit listening to a
minstrel's lay, while contentment possesses the people . 78
The lay of the Wooden Horse is sung at Odysseus' request:
'I admire you above all men, Demodokos. Apollo must have been your
teacher, or the Muses, daughters of Zeus. I have never heard such a rendering
of the misfortunes of the Achseans, just as though you had been there. Now
give us another — the story of the Wooden Horse. . . .' After starting from
God the minstrel began the lay . 78
‘Starting from God’ — the same formula, back in the heroic age.
And the words seem to imply that it was not merely a formal
opening but something distinct from the lay.
At die beginning of the Odyssey Telemachos is entertaining a
stranger — really Athena in disguise. He is anxious to enquire
after his lost father, but the presence of the suitors embarrasses
him. He gets his chance after supper. ~
When the meal was over, the suitors gave their minds to singing and
dancing, for these are the delights of a feast. Phemios, the minstrel, who
sang to them against his will, took his lyre from the herald and struck up
a lay. Then Tclcmachos spoke to Athena, their heads close together, so that
they should not be overheard . 86
At the end of the conversation the stranger vanishes into the
air, and Telemachos returns to the suitors, who are sitting
quiedy listening to the minstrel as he sings of the return of the
78 01 8. 72-82, 485-95. 77 Ol 1. 325-7. 78 oi 9. 3-10.
70 Oi. 8. 487-99. 86 Oi. 1. 150-7.
XV THE RITUAL ORIGINS OF GREEK EPIC 495
Achjeans . 81 But we were told before that they gave their minds
to singing and dancing. How did this lay begin? It looks as
though die solo — the heroic lay-— had been preceded by a
choral ode.
This is the sequence we found at Delos. There, a hymn to
the gods followed by an epic recital; here, a choral ode followed
by a heroic lay. The epics were intrusive at Delos — they had
been brought there ready-made; but this heroic lay is in its
original setting — the palace' of an Achaean chief. It cannot be
intrusive; it must have grown out of the ode. Taken together
with its choral prelude, it follows the same pattern as the
Hesiodic Theogony and Catalogue and the Delian hymn.
Songs after supper did not cease to be sung after they had
given rise to the heroic lay. They were an established conven-
tion among the nobility throughout the historical period.
This convention is worth examining, because, since it was
maintained in its original setting, it may reveal features which
the art of minstrelsy discarded. The best-known examples are
the Attic drinking-songs.
After supper, when the wine was brought in, the company
sang a paean to Apollo while libations were poured to the gods,
the heroes, and Zeus the Saviour . 82 Then the wine was served.
A loving-cup was passed round together with a branch of
myrtle or laurel (ahakos). Each guest took the cup and the
81 Oi. 1. 325-7.
82 A.A. 257-8, X.Sym. 2. 1, PI. Sym, 176a, Sch. PI. 916a. 28.
496 STUDIES IN ANCIENT GREEK SOCIETY XV
branch in turn, and branch in hand he improvised a stanza,
usually on some secular theme, political or aphoristic.”
In this branch we recognise the rod plucked by die Muses
for Hesiod 84 and the staff of die Homeric rhapsodes. In die
pzean we recognise the epic prelude (proofmiott 1) and in the im-
provised stanzas the heroic lay.
This custom was no doubt indigenous in Attica, but it was
developed under influences from Ionia. The Ionian convention
owed its artistic form largely to Pythermos of Teos, who lived
early in the sixth century. One fragment of his survives,
showing that his stanza was the same as the Attic . 86 The Attic
was as follows:
<pIATO0’ 'AppdSi', ou t! ttco •rffivrjKaj,
tiruir u“iru
vi^crois S' tv pctK&pcov at tpaaiv clvoa,
-U-UU- U“U“U
Tvoc Trep troSdJKris 'AxiAeOj
UU“U- “UlT
TuSeIS^v ft <paaiv MMv AiopriS ia.
~UU~U~U~ ~uinr
The resemblance to the Alcaic is unmistakable. Again we
have a stanza of four phrases, the first two being identical.
The opening is in fact the same as the Alcaic, except that the
order of the two figures is reversed and the trochee precedes the
dactyl. The two stanzas belong to the same convention. And
here it may be added that Alkaios himself was celebrated for
his drinking-songs. Many, perhaps all, of his monodies belong
83 PI. Go. 451c, Ath. 694a, Ar. Ntt. 1364 sch., PIu. M. 615b. It would
be interesting to follow up the history of this convention, which still
survives in Greece and is evidently very old. In Anglo-Saxon England, when
the villagers met to drink in the evening, everyone took his turn at singing
to the harp (Bede Eccl. Hist. 4. 24). In Ireland, early in the last century, the
Limerick school of poets used to forgather in the same way, improvising
stanzas on a fixed pattern which one has only to hear to recognise the origin
of the limerick (Dinncen FM). In Cambridge during my student days
similar symposia were held annually, but improvisation bred impropriety,
and they have been banned.
84 Hes. Th. 30-1. 86 Diehl 2. 60.
XV THE RITUAL ORIGINS OF GREEK EPIC 497
to this class, and several of those expressly assigned to it are
in the Alcaic stanza.
In the Attic symposium the secular part of the programme
was rendered by the whole company, singing a round of solos.
FIG. 77. Dancing girl: Attic cup
In the symposium of the heroic age it took the form of a single
extended solo sung by a professional minstrel. That is the main
difference between them, and it is readily explained by the
special conditions of the heroic monarchy. The minstrels were
court retainers, patronised and encouraged by the kings, and
their art was a specialised occupation. The rude tribal chiefs
from whom these kings were descended had joined in the
singing themselves. Even in Homer the memory of this state of
things is not entirely effaced. Achilles whiles away the tedious
Hh
498 STUDIES IN ANCIENT GREEK SOCIETY XV
hours by singing to his lyre of ‘the glories of men *. 88 This does
not mean he was a professional minstrel. In the rough country
he came from the art was less specialised, more widely diffused,
than it was at Mycens.
One more complication remains to be unravelled, and then our
argument will be complete. Some of Alkaios’ drinking-songs are
in the Sapphic stanza . 87 Did Sappho write drinking-songs
tool
The idea of a feminine symposium did not find favour with
Victorian scholars, who dismissed it as a slur on the fair name
of Greek womanhood . 88 They did not pause to reflect why
there should have been any impropriety in women refreshing
themselves after the same fashion as their husbands in a
country where the worship of the wine-god was mainly their
business and water is scarce. It is true that no such custom
is recorded in democratic Athens, but Athens was not Greece,
and democratic Athens was notorious for its peculiar attitude
to women. The ancient authorities do not seem to have felt
any qualms. Praxilla, a lady of Sikyon, is mentioned by them
as a composer of drinking-songs, and so is Sappho . 89
Sappho was in charge of a finishing school for the young
ladies of Lesbos. I call it a finishing school, but it might be
more aptly described as an initiation school, which is the same
institution in a more primitive form. It was in fact a female
cult society like the Spartan agcla for which Alkman composed
his parthfaeion .
The great moments in the life of this little coterie were the
days when one of the pupils left to be married, with a wedding
song composed for her by Sappho . 80 The girls participated in
the civic cycle of women’s festivals, and one of the most im-
portant of these was the feast of Adonis, for which Sappho
composed dirges . 91 We are not informed how these young
ladies spent their evenings, but they must have had their
88 11 . 9. 186-9. 87 Ale. 77 - 8 . 85, 92.
88 Reitzenstein 18-9, H. W. Smyth cv. It was the custom in Illyria
for women to be present at men’s drinking-parties: Ael. VH. 3. 15.
That a similar custom had prevailed in prehistoric Greece may be inferred
from Od. 4. 219-34, A.Ag. 254-8. And of course it survived at Athens in
respect of slave girls and courtesans: PI. Smp. 176c.
89 Ar. V. 1240 sch. eo Sa. 115-33, cf* 96, 98.
°i Sa. 21, 107.
XV - THE RITUAL ORIGINS OF GREEK EPIC 499
private vespers, at which they sang Sappho’s hymns, and so it
cannot be regarded as improbable that they sang songs after
supper.
At the feast of Adonis in Samos the girls used to hold
drinking parties, at which they propounded riddles . 88 It is not
likely that this feature of the cult of Adonis was peculiar to
Samos, because it is recorded of other cults. In Boeotia, at the
Agriania, the women used to go out in search of the lost
Dionysus, and after supper they spent the evening in asking
riddles . 83
The riddle, which is as universal as it is ancient, was in
origin a vehicle for catechism in initiatory secrets . 84 In Greece,
■ as in most parts of Europe, it degenerated eventually into a
children’s game of forfeits, but its magical import was re-
membered in the story of Kalchas, the soothsayer who died of
vexation after failing to answer a riddle put to him by a rival . 86
Stories of this type are a commonplace of Indo-European
mythology . 86 In Greece the riddle preserved its metrical form,
which means that it was sung, or had been.
At these festivals, then, we have to envisage groups of
women, sitting perhaps out of doors like the Minoan ladies on
the fresco and conducting extempore a running musical
catechism on themes appropriate to the occasion. The picture
is incomplete, but the parallel with the men’s supper parties
is obvious. Both go back ultimately to the undulating series
of solo and refrain circling endlessly in the twilight round the
clan camp-fire.
At the outset of this study we saw how mimetic impersona-
tions of the activities of the clan totem passed into dramatic
dances commemorating the achievements of the'’ clan ancestors,
whose magical energies were thereby evoked to fertilise the
sources of food-supply. We also saw how, with the growth of
class inequality, these ancestral spirits became gods. Even the
Greek gods preserved ancestral ties with their worshippers as
progenitors of ruling clans. ‘The race of gods’, says Pindar,
‘and the race of men are one .’ 87 But in general, with the
88 Ath. 451b. » 3 PIu. M. 717a.
84 Pauly- Wissowa j.v. Ratsel, Chadwick GL 3. 152-31 834-6.
86 Str. 642-3. 66 Chadwick GL 1. 474. » 7 Pi. N. 6. 1.
500 STUDIES IN ANCIENT GREEK SOCIETY XV
consolidation of class society, the gods came to appear as a race
apart, showing a parental regard for their worshippers but en-
joying the privilege of everlasting life.
In the same way we followed the development of the
totemic rite into a sacrifice, a feast shared with the god and
attended with dance and song. This was the genesis of the
choral ode, in which the human company, having enjoyed
their meal, began by praising their gods and went on to recall
the traditions of their heroic ancestors. At first, in conformity
with the matriarchal structure of society t the parts had been
taken by women, who sang the praises of goddesses and
heroines; but in later times, with the extension of warfare and
the accumulation of personal property in the hands of military
chiefs, many of whom were patriarchal newcomers from the
north, there arose a new type of choral ode — martial, masculine,
personal, secular. The 'glories of women faded. The ‘glories of
gods' retained their pride of place but were curtailed and recast
to fit their new setting. Interest was concentrated . in the
‘glories of men’ — the men actually present, listening to the
minstrel. The male poet had dismissed his chorus of dancers,
and it only remained for him to discard his lyre.
XVI
HOMERIC ARCHEOLOGY AND LINGUISTICS
I. Datable Elements
The Attic tragedians presented their heroes on the stage in a
more or less contemporary setting, drawing on the ideas and
customs of their own time without regard for historical con-
sistency. The epic tradition was quite different. Like heroic
poetry in general, it was consciously archaistic. Mycenean
civilisation is described in the poems as though it were still
flourishing, while everything that had happened since is
studiously ignored. There are no Dorians in the Peloponnese,
no lonians in Asia Minor; weapons are of bronze; gold and
silver are plentiful. These poets lived in the past. There are of
course discrepancies. From incidental allusions, let fall in-
advertently, we see that they were well acquainted with the use
of iron, and we have detected beneath the surface a good deal of
confusion regarding the status of women (pp. 416-30). But
the general accuracy of their antiquarian knowledge has been
confirmed by archaeology.
Homeric archaeology is a comparative study. Its object is to
interpret the poems in the light of the excavated remains and
the remains in the light of the poems. There are elements in the
poems — descriptions of material objects and social usages —
which have been dated by archaeologists to definite periods,
early or late, from the fifteenth century to the seventh. They
have been discussed many times. Here I shall select only the
clearest instances, with the object of elucidating certain
principles of Homeric criticism.
In Book XI of the Iliad Patroklos looks in at Nestor's tent
and sees on the table a cup:
It was studded with gold nails; it had four handles, with two gold doves
on either side and two stems beneath them . 1
1 II. 11. 632-5.
502 STUDIES IN ANCIENT GREEK SOCIETY XVI
The last detail baffled Homeric scholars until there turned up
in the fourth Shaft Grave at Mycenas a chalice with curiously-
shaped handles answering closely to this description . 9
In Book X Meriones lends Odysseus his
helmet:
It was made of leather, held together on the inside
by string and on the outside by boar’s teeth studded
all over it. 2 3
This puzzle was solved by representations
hg. 78. Gold cup of helmets in Mycenean art, supplemented by
from tie Fourth the discovery in tombs of pieces of boar’s
Shaft Grave tusk cut to shape and perforated at one end for
attachment. The pieces were fitted into the leather cap and
threaded together on the inside . 4
These two objects have been dated to the fifteenth century,
thus confirming the accepted view that the Homeric tradition
goes back into the Mycenean age. But
they do not prove that the passages in
which they are described are as old as
that, because inheroicpoetry such antique
descriptions are commonly preserved as
traditional themes, told and re-told for
generations. Only the content of the
passages is dated.
In Book XI Agamemnon puts on his
cuirass, which was apresent from Cyprus:
It had ten bands of cyanus, twelve of gold,
twenty of tin, and three cyanus serpents circled
like rainbows up to the neck.®
The snake was not used for decorative
purposes at Mycenae, but it was common
in Phoenicia and in early Greek art of the oriental style , 5 This
cuirass cannot be much older than the seventh century.
In Book XVn one of the Trojan allies, Euphorbos, is de-
scribed as wearing his hair in ‘plaits bound with gold and
silver '. 7 Mycenean men did not wear plaits, but they were
2 Nilsson HM 1 37-8: fig. 78. 3 Zl. 10. 261-5. 4 Nilsson HM 138: fig. 79.
8 & 11. 19-28. « Nilsson HM 125-6. 7 IL 17. 52.
XVI ARCHAEOLOGY AND LINGUISTICS 503
fashionable with both sexes in the sixth century . 8 Euphorbos*
plaits are no older than Agamemnon’s cuirass.
While the form of words in which an object is described
may be later than the object, it cannot be earlier. If the
cuirass and the coiffure belong to the early historical period, so
do the passages themselves, and, since there must be others
equally late which have not yet been detected, we conclude
that the poems were still expanding in the seventh century.
For fixing the chronology of the poems as a whole isolated
passages like these are not enough. We must look for elements
so - deeply imbedded or so pervasive that they cannot be
explained as mere accretions. There are several of these, but
most of them are still controversial. I shall confine myself to
two — the mode of burial, and Helen.
2. The Mode oj Burial
The Mycenean princes buried their dead. In the poems they
are cremated. In later Greece the two practices existed side by
side. This is one of the most-discussed contradictions of
Homeric archaeology. Let me begin with some remarks on burial
customs in general.
Wherever interment is the rule, it has been, and still is,
customary to deposit beside the corpse pots, tools, weapons,
utensils of all kinds. This is explained by most authorities as a
means of equipping the deceased for a future life, and the same
reason is given in many instances by the peoples themselves. It
is well known, however, that new motives are constantly in-
vented to justify the continuance of practices that have ceased to
serve their original purpose. In .the present case the motive
alleged involves serious difficulties. Not all the deposits have a
utilitarian value. Some of them — figurines, phalli, amulets —
are clearly magical. Moreover, the pots are often broken de-
liberately before they are thrown in.® These at least were not
intended for use in the hereafter. It is much more likely, as
Karsten has argued, that they were broken in order to release
8 Nilsson HM 127-30.
9 Karsten 244-5, 246, 251-3, Roscoe B (1923) 147. The practice
still survives in Greece: Polices BVFR.
504 STUDIES IN ANCIENT GREEK. SO CIETY XVI
the magic inside them; and in that case the same may be pre-
sumed of the other articles as well . 1 0 As the dead man's personal
effects, they retain something of his life, and so will be specially
potent in restoring him. On this interpretation the practice
falls into line with other burial customs equally widespread —
placing the body in the antenatal posture, painting the bones
with red ochre, scattering grain or leaves beside the corpse,
planting flowers on the grave. The ceremony of interment is
only a specialised rite of initiation. Its object is the renewal of
life.
So with cremation. Since birth is death and death is birth,
regeneration requires mortification. The old Adam must die
before the new man can be raised up in him. One of the com-
monest initiation rites is the ordeal by fire, which is merely
regeneration in its purificatory aspect. And that is its signific-
ance in the disposal of the dead. Interment and cremation are
simply the positive and negative aspects of the same principle
which with the disintegration of the principle have become
distinct.
It has been suggested that cremation implies a more materi-
alistic outlook than interment . 11 This too is a misconception.
The belief that a corpse can be restored to life by contact with
broken crockery is a typical piece of crude primitive materialism.
In the ideology of cremation the dead man survives merely as a
disembodied spirit. The Homeric concept of the soul as a
ghost or shadow of the living person was a step on the road to
Orphic mysticism, in which the soul was treated as immaterial
and immortal.
Approaching the problem from this point of view, we find
that the gulf between Mycenean practice and Homeric tradi-
tion is neither so wide nor so deep as has been supposed.
Many Mycenean tombs contain traces of fire, and an un-
rifled tomb at Dendra has revealed what the procedure was.
The corpse was interred, but the personal effects were burnt in
a shallow trench beside the grave. A second trench contained
the charred remains of victims, human and animal, which had
been slaughtered at the funeral . 12 At the funeral of Patroklos
Achilles slaughters a number of dogs and horses together
10 Karsten 244-5. 11 Lorimer PU 177. 12 Nilsson HM 155.
XVI ARCHEOLOGY AND LINGUISTICS 505
with twelve Trojan captives and flings their bodies on to the
fire he has lit under the corpse. 18 Here the corpse itself is
burnt. This incident is described as something out-of-the-way
and horrible. The normal Homeric burial is much simpler.
After a sacrifice of animal victims the corpse is burnt, the
bones and ashes deposited in a box, and a mound heaped over
them with a gravestone on the top. 14 Is this cremation or in-
terment? It is surely both.
Simple interment was current in the eighth and seventh
centuries, 16 so the Homeric poets must have been familiar
with it. But they do not mention it. This must mean that they
were adhering to what they believed to have been the heroic
practice. As we have just seen, their tradition was inaccurate.
In Mycenean times the body had been interred and the be-
longings burnt; in the poems the body is first burnt and then
interred. The Homeric procedure is simpler, and surely it
must be derived from the Mycenean: why else should the
ashes be buried in the ground? If this was the rule among the
AE’olic and Ionic nobility descended from the dynasties of
Mycenae and Pylos, that would explain why the minstrels
accepted it as valid for the heroic age; and, if we ask what
induced these Imigris to depart from die practice of their fore-
fathers , the answer is that in the reduced circumstances of their
new homes they could not afford it.
This conclusion would carry little weight if it stood alone,
because I am not an archaeologist; but it is close to the result
reached by Lorimer, who even considers it possible that the
Achaeans had been cremating their dead in Greece itself from
an indefinitely remote past, only modifying the practice under
Mycenean influence. 16 If they had simply burnt the body to ashes
without inhuming the remains, there would be nothing for the
archaeologist to recover.
3. Helen
Helen, whose face it was that launched the thousand ships,
is a myth — a myth of the eternal fragility of woman's beauty:
18 II. 23. 164-9.
14 II. 24. 788-801, 6. 418-9, Oi. 1. 291, 2. 222, II. 21. 320-1, 23.
91, 239, 252-3, Od. 24. 65-84.
16 Lorimer PU 170-1. 18 lb. 176.
506 studies in ancient Greek society xvi
Brightness falls from the air,
Queens have died young and fair,
Dust hath closed Helen’s eye.
I shall try to show that a methodical enquiry into the genesis of
this myth enhances our admiration for the poets who created it.
Her mythical origin is now generally recognised. 17 Begotten
by Zeus disguised as a swan, bom from the egg laid by the
Carian ‘woman’ (p. 429), she is akin both to the Caro-
Lelegian Artemis of the Marshes, who was represented in cult
as a waterfowl, 18 and to the Phoenician Aphrodite-Astarte, who
was hatched from an egg that fell from the moon. 18
17 Nilsson MOGM 74-5, 170-5.
1® Harrison T 1x4, Imhoof-Blumer CTK 99.
18 Hyg. F. 197- Helen's egg too was said to have fallen from the moon:
Ath, 57F, Plu. M. 637b.
XVI ARCHAEOLOGY AND LINGUISTICS 507
The story of her elopement was told in the Kypria, one of
the lost Homeric epics known to us only in epitome. While
Paris was at Sparta on a visit to Mcnelaos, his host was called
away to Crete, and three days later Paris landed with Helen at
Troy. The story is mentioned only once in the Iliad, and then
incidentally, where Hecuba selects a robe for Adiena:
She went into her bedchamber, where she stored her embroidered robes,
woven by women of Sidon whom Paris had brought back with him on the
voyage on which he had fetched Helen. 2 ®
As Herodotus observed, this passage contradicts the Kypria . 21
Why did Paris return from Sparta to Troy by way of the
capital of Phoenicia?
The wanderings of Mcnelaos after the fall of Troy are nar-
rated in the Odyssey."- After quarrelling with Agamemnon, he
embarked without him, being anxious to cross the AEgean
before winter. Off Lesbos he came up with Nestor, also hur-
rying home, and die two sailed together as far as Sounion.
There Menelaos was delayed by the death of his helmsman.
Resuming his journey, he was making Cape Malea, when a
storm carried him off to Crete and Egypt. It was seven years
before he got home. He visited Cyprus, Phoenicia, Ethiopia,
Egypt, and Libya. While in Egypt, he neglected to perform a
sacrifice demanded of him by the gods — its nature is not
stated — and in consequence was held up by bad weather in the
island of Pharos. There he met Proteus, the Old Man of the
Sea, who foretold his destiny:
It is not your fate, Menelaos, to die in Argos. You shall be taken by the
immortals to the fields of Elysium, where Rhadamanthys is and life is
easiest for man, a country without snow, rain, or storm, cooled perpetually
by the west wind sent by Ocean to refresh mankind. This fate has been
given to you because you are Helen's husband and son-in-law of Zeus. 23
After that Menelaos returned to Egypt, performed the sacri-
fice, and made his way back to Sparta, where we meet him
telling the story to Telemachos, with Helen busy over her
work-basket in the firelight.
The remarkable thing about this story is that, with the
2 « II. 6. 288-92.
23 Hdt. 2. x 17, cf. A. A. 696. The passage in the Kypria was subsequently
rewritten so as to conform to the Iliad: Prod. Cir. p. 103, cf. Alien HOT 151.
22 Oi . 3. 130-69, 276-302, 4. 351-586. 23 01 4. 561-9.
508 STUDIES IN ANCIENT GREEK SOCIETY XVI
single exception of the passage just , quoted, it ignores Helen.
It was for her that ‘the princes orgulous, their high blood
chaf'd', had vowed to ransack Troy; yet we are left to assume
that she was restored to her husband after ten years of blood-
shed and that she was his constant companion during his seven
years' seafaring. All we learn of her part in his adventures is
what we can infer from two passing allusions. Her work-
basket is a present from the wife of Polybos, king of the
Egyptian Thebes, and she also possesses a drug, an antidote to
grief, which had been given her in Egypt by Polydamna, the
wife of Thon . 24
Helen went to Phoenicia before the war; she returned from
Egypt after it. Had she ever been in Troy?
Stesichoros said no. Only her wraith went to Troy. The war
was fought for a phantom. In an earlier poem Stesichoros had
accepted the Homeric version; then he was struck blind and
wrote his famous palinode: ‘That story is untrue; thou didst
not set foot on shipboard nor go to the towers of Troy .' 26
The recantation has sometimes been accepted at its face value,
but it is hard to believe that a Greek poet would have dared
to defy the Iliad unless he had some alternative authority to
rely on. And it appears that Stesichoros had, for according to
Tzetzes the idea of a phantom Helen had already been put
forward by Hesiod . 28
The problem is discussed by Herodotus, who with all his
shortcomings as a historian was a shrewd literary critic. He
refers to it in his account of the Egyptian kings:
Pheros was succeeded by a native of Memphis, whose Greek name was
Proteus. In Memphis there is a fine, well-appointed tcmenos consecrated to
this king. It stands to the south of the temple of Hephaistos in a Phoenician
settlement, the whole quarter being known as the Tyrian Camp. Within
the timmos is a shrine of Aphrodite the Stranger, whom I take to be Helen,
daughter of Tyndareos, because I know the story of her visit to Proteus, and
there is no other shrine of Aphrodite with this title. 2 ?
s* OH. 4. 125-32, 220-30.
25 Stcs. II. The Homcridai had a tradition that Helen visited Homer
in a dream and told him to compose a poem on 'the expedition to Troy’:
Iso. Hel. 64-5. For other stories of the same sort see Paus. 9. 23. 3, Pi.
Plio 6oe, Plu. M. 543 a *
sc Hcs. fr. 266=Lyc. 822 sch. 27 Hdt. 2. 12-13.
XVI ARCHAEOLOGY AND LINGUISTICS 509
He goes on to repeat the story as he had heard it from the
priests of the shrine. After the elopement Paris embarked for
Troy, but he was carried out of his course and cast up in
Egypt. There some of his servants informed the priests of
what he had done to Menelaos, and Thonis, warden of the
Nile mouth, reported the matter to Proteus. Paris was then
arrested, and after taking Helen into his own custody Proteus
ordered him to leave the country in three days.
That is how Helen came to Proteus, according to the priests, and I think
Homer must have been acquainted with this version and deliberately sup-
pressed it as less suited to the spirit of epic. It is clear that he knew of it,
because in the one passage in which he refers to the wanderings of Paris he
tells how after being driven out of his course he travelled with Helen to
Sid on in Phoenicia. 28
Herodotus then quotes the passage about the robe, and after
contrasting it with the version in the Kypria he relates how
Helen was recovered:
On his arrival in Egypt Menelaos went upstream to Memphis, where,
after recounting his adventures, he was hospitably entertained and Helen
was restored to him together with the property which Par is had stolen at the
same time. Then, although the Egyptians had treated him so handsomely,
he did them a bad turn. Being detained for a long rime by storms, he took
two Egyptian children and sacrificed them. When the crime became known,
die angry people raised a hue and cry, and Menelaos took ship and fled to
Libya. 28
The recurrent names — Proteus, Pharos or Pheros, Thon or
Thonis — show that the Herodotean version is connected in
some way with the Homeric, but on the main issue it con-
tradicts it. Helen never went to Troy, only to Egypt. How old
was this story of Helen’s stay in Egypt?
The island of Pharos lies just outside the Delta. In the
Odyssey it is described as being a whole day’s journey from the
coast . 30 Such a miscalculation would have been impossible
after 650 B.C., when the Greeks established a trading station
in the Delta at Naukratis, and they must have been familiar
with the approaches to the Nile for a good many years before
that. Further, as Lorimer and Nilsson have pointed out, the
allusion to the Egyptian Thebes implies that it was the royal
28 Hdt. 2. 116. 28 Hdt. z. 1 19. 80 Od. 4. 354-7.
510 STUDIES IN ANCIENT GREEK SOCIETY XVI
capital. The city was completely destroyed by Ashurbanipal in
663 b.c., and it had not been die capital since the fifteenth
century (p. 3 80). 81
The antiquity of the Herodotean version is confirmed by
another point. It supplies what is missing in the Homeric. The
casual allusion in die Odyssey to the sacrifice is explained by.
what Herodotus was told at Memphis, which shows that
Menelaos resorted to a rite of the same nature as his brother
had done in a similar predicament. I refer to the sacrifice of
Iphigeneia.
I Herodotus' informants were the priests of Aphrodite in the
Tyrian Camp. The tradition was therefore of Phoenician origin.
And here we must notice yet another version, preserved by
Dracontius. 32 Paris met Helen in a temple of Aphrodite in
Cyprus, and embarked with her for Troy just as Menelaos
arrived from Crete. Dracontius was a native of Carthage, and
so he too was a Phoenician.
.Herodotus and Dracontius agree in suggesting that Helen
belonged properly to the Levant, where she was a hypostasis of
Aphrodite-Astarte. This is what lies behind the Homeric
allusion to the voyage to Sidon. It must have been the
Homeridai who domiciled her at Sparta and sent her to Troy.
They tried to forget her Levantine origin, but it slipped out
~ inadvertently, to be eliminated later in the Kypria. It was re-
membered by the Hesiodic school, who tried to restore order
by inventing a distinction between the real Helen and the
phantom.
It may be objected that there were no Phoenician contacts
with the Aegean before the ninth century. This is true. It is
proved by Phoenician silver-ware excavated on Greek sites,
copied from Egyptian and Assyrian models. These belong to
the eighth century. Rather earlier, we have Egyptian scarabs
and figurines, which were probably traded in Phoenician
vessels. It was doubtless in this period that Aphrodite became
known as the Cyprian and the Cytherean. 88 Cyprus was settled
by Phoenicians about this time, and Kythera was one of their
AEgean trading-stations. 34 Nilsson accordingly holds that
si Lorimer HUP 153, Nilsson HM 157-8. 32 Drac. Rapt. Hel.
33 II . 5. 330, 8. 288, 18. 193, cf. Paus. 3. 23. 1. 34 Hdc. 1. 105.
XVI ARCHEOLOGY AND LINGUISTICS 511
the Phoenicians cannot have entered the yEgean before the
tenth century at earliest — 200 years after the traditional
date of the Trojan War . 36
The stories of the Phoenicians told in the Odyssey imply the
presence of Phoenician seafarers in .Egean waters, ana must
36 Nilsson HM 134.
512 STUDIES IN ANCIENT GREEK SOCIETY XVI
therefore be assigned to the tenth century or later. 88 But the
voyages of Paris and Menelaos imply the presence ofAEgean sea-
farers in Phoenician waters. They hark back to the time when
the Phoenicians were not yet in control of the Levant — when the
Peoples of the Sea were pouring down into Syria and Palestine
and harrying the Delta. Among these were the Achaeans, who,
as we learn from Hittite documents, had established themselves
in Cyprus as early as 1240 B.c. 87 Pheros belongs to the same
period; for he is identified by Egyptologists as a king of the
XIXth Dynasty. 88 Can the Phoenician Aphrodite be traced so
far back? It seems she can.
In Chapter XI, discussing the origin of Kadmos, we had
occasion to mention the excavations recently carried out in
northern Syria, which have revealed contacts with Minoan
Crete and Mycenean Greece and suggest that some elements of
Middle Minoan culture may have been derived from this
area (p. 376). If so, there was a two-way movement — from
Syria to Crete in the Middle Minoan period and from Mycenae
to Syria after the fall of Knossos (p. 374). The first ex-
plains why Kadmos, who brought Demeter from Crete to
Greece (p. 124) was regarded as a Phoenician; the second sug-
gests that the myth of Helen originated among the Achaean
sea-raiders of the Levant.
As a dove-goddess, Aphrodite-Astarte is descended from the
Kupapa of the Hittite hieroglyphs, the goddess of Carche-
mish; 89 and this Kupapa, with her consort Sandas, emanates
from S.W. Anatolia, where we meet her as Kybebe. 40 She
must also be related in some way to the Minoan dove-goddess
(p. 251), one of the parents of the Greek Aphrodite. Thus,
while the historical Aphrodite owed two of her titles to the
Phoenician goddess introduced in the ninth and eighth cen-
turies, this goddess herself had been fashioned under the. in-
fluence of AEgean settlers in Syria. All this goes to show that
the story of Helen told by the priests of Aphrodite the
Stranger in the Tyrian Camp was an independent tradition
which the Phoenicians had once shared with the Achasans.
aa Ol 13. 271-86, 14. 288-91, 15. 415-84, cf. ll. 23. 740-7.
3 ? Cavaignac 95. 88 Wainwright 75.
8» Cavaignac 168. . *0 Hde. 5. 102, Hsch. Kvp^pii.
XVI
ARCHAEOLOGY AND LINGUISTICS
513
If Helen dissolves into a goddess, what becomes of Menelaos?
He is more substantial than she is, yet even he is not entirely,
of this world. He was destined, as we have seen, for Elysium,
and this unique privilege was conferred on him ‘because he
was the son-in-law of Zeus'. 41 Overlooking the
Eurotas, about four miles from Sparta, stands
the Mycenean site of Therapne. Here there was
a sanctuary in which he and Helen were said to
lie buried. 42 This does not contradict the legend
of his immortality; rather, it explains it, because
he was worshipped here with Helen as a god. 43
Was he a priest-king ruling by right of marriage
to the local Helen-Aphrodite?
This is only a conjecture, but it is supported
by an' analogy. The Kinyradai, priest-kings of
Cyprus, claimed descent in the male line from
an Achaean chief, Teukros, who had married a
daughter of Kinyras, the priest of Aphrodite. 44
Their Achaean connections were remembered in
Homer: it was Kinyras who presented Agamemnon
with his cuirass (p. 502). Their palace was at
Paphos, Aphrodite's dwelling-place, 46 and here
times stood one of her greatest temples. The
lay in the precincts, and the priesthood was
fig. 81.
Dove-headed
Aphrodite:
Cyprian
terracotta
in historical
royal tombs
a prerogative
of the family. 46 That they were regarded in some sense as
consorts of the goddess is suggested by the tradition that Kypros,
the eponym of the island, was a child of hers by Kinyras. 47
This cult had reached Cyprus from northern Syria. At
Lebanon there was a shrine of Aphrodite built by Kinyras. 48
At Byblos, which was sacred to Adonis, the beloved of
Aphrodite, there was a palace of Kinyras. 49 Kinyras is even
41 Kadmos and Rhadamanthys were also sent to the Elysian Fields, which
are the Phoenician ‘Fields of El’: Schaeffer 61.
48 Paus. 3. 19. 9, Hdt. 6. 61. 3. 48 Iso. Hel. 63.
44 Paus. 1. 3. 2, Pi. P. 2. 15-7. 46 Od. 8. 362-3.
46 Tac. H. 2. 3, Paus. 2. 29. 4, Ptol. Meg. 1 =FHG. 3. 66, Sttpp. Ep. Gr.
6. 820.
47 St. B. s.v. Ktinpos, Philosteph. n =FHG. 3. 30.
48 Luc. DSyr. 9, Str. 755, cf. Hdt. 1. 105. 3.
49 Str. 755, Luc. DSyr. 6; Hooke MR 82-3.
Ii
514 STUDIES IN ANCIENT GREEK SOCIETY XVI‘
described as a king of Assyria . 60 All this hangs together. In myth,
Ishtar and Tammuz, the Babylonian mother-goddess and her
male partner, were transmitted through Syria and Cyprus to
Greece as Aphrodite and Adonis; in ritual, the priest-king
appears in Syria and Cyprus as Kinyras, in Sparta as Menelaos.
If this analogy is sound, it explains the central event in the
life-story of our heroine — her elopement. That die rape of
Helen is founded on ritual is generally agreed. There are many
parallels, but the closest is in the cult we have just examined.
Like the same goddess in Syria and Babylon, Aphrodite was
served at Paphos by sacred harlots . 61 Herodotus says that the
Cyprian form of the institution was similar to the Babylonian,
which he describes in detail:
Every woman is obliged once in her life to seat herself at the temple of
Aphrodite and give herself to a stranger. They sit all together in the pre-
cinct with crowns of cord on their heads. Women are constantly coming
and going, and passage ways are roped off through the crowd for the men to
walk in and take their choice. Once she has taken her place, a woman never
goes home till a stranger has had intercourse with her after putting money in
her lap, and as he does so he says: ‘I call on you in the name of Mylitta.’
This is the Assyrian for Aphrodite. 62
This custom underlies a story of the daughters of Kinyras.
Aphrodite obliged them to cohabit with strangers, and then
they fled to Egypt . 63 So did Helen.
If the old oriental Helen, who never went to Troy, survived
in the Hesiodic tradition despite Homeric influence, the
Homeric Helen must have been fashioned in Aiolis or Ionia
after the coming of the Dorians. The woman whose beauty
never ceased to take men’s breath away through ten years of
blood and tears was a creation of the poets who immortalised her:
The old men, elders of the people, were sitting at the Scaean Gate. They
were past fighting but good talkers, like cicadas murmuring softly in the
woodland trees, and, as they saw Helen pass, one said to another: 'Small
blame to the Trojans and Achaeans for enduring so much for such a woman.
She is terribly like the immortal goddesses to look at. Still, let her go, or
she will be the death of us and our children.' 64
60 Hyg. F. 58, 242, 270.
61 Hdt. 1. 199. 5, Clem. Pr. 2. 13, Amob. Adv. Nat. 5. 19.
ss Hdt. 1. 199.
53 Apld. 3. 14. 3; Frazer GB-AAO 36-41.
64 II. 3. 146-60.
XVI ' ARCHAEOLOGY AND LINGUISTICS 5 I 5
Thus, while the roots of the Iliad and Odyssey lie far back in
the Mycenean age, the poems as a whole seem to have taken
shape in Asia Minor during the tenth and ninth centuries, and
they were still expanding in the seventh. This archeological
result accords with the traditional date of Homer, who in the
opinion of Thucydides lived 'long after the Trojan War ,
while Herodotus places him ‘not more than 400 years ago’, i.e.
about 950 B.c. 85
4. The Epic Dialect
The language of the poems differs from all the known
dialects of Greek, spoken and literary. It is on the face of it a
mixed dialect — mainly Moliz and Ionic, with a good deal of
Arcado-Cyprian and a touch here and there of Attic. These
poems are the earliest Greek documents we possess. Apart from,
some lyric fragments and a few short inscriptions of the seventh
century, we have nothing else older than the sixth. The genesis
of the epic dialect has to be reconstructed from the remains of a
much later period. A simple illustration will indicate the nature
of the problem. The Ionic for ‘house' was oTkoj. In >Eolic and
Arcado-Cyprian it was foikos. The latter form was the older,
corresponding to the Latin views and the English -wick Our
text of Homer gives always oIkos, without the digamma, but
the word is usually so placed in the verse that the digamma is
required by the metre, showing that the Homeric form had
once been f°ikos. Was this JBolic, Arcado-Cyprian, old
Ionic, or simply proto-Greek?
Three hypotheses have been advanced to account for the
mixture of 4 Eolic and Ionic. They have been discussed many
times and can be dealt with very briefly.
Wilamowitz and Allen maintained that Homeric Greek was
founded on the local dialect of the middle region of the
Asiatic coast, where 4 Eolic and Ionic overlap. 86 This region
includes Smyrna and Chios, both of which claimed to be
Homer’s birthplace. The dialect survives in inscriptions.
It is Ionic with an admixture of aEoKc, but its correspondences
with Homeric Greek are not dose enough to establish a direct
68 Th. 1. 3. 3, Hdt. z. 53. z.
88 Wilamowitz IW 61, Allen HOT 98-109.
516 studies in ancient Greek society xvi
connection. In particular, some of its characteristic AEolicisms
are absent from Homeric Greek . 67 The epic medium may have
originated here, but this evidence does not prove it.
Meyer contended that Homeric Greek represents the parent
of iEoIic and Ionic. This view rests on the assumption that
these dialects were only differentiated after the colonisation
of the Asiatic coast . 68 But Ionic is closer to Attic than to
AEolic. It must have separated from AEolic before it separated
from Attic, and its separation from Attic cannot be put later
than the Ionian migration. Consequently, when the Greeks
settled on the Asiatic coast, AEolic and Ionic were already
distinct.
These two hypotheses start from the assumption that
Homeric Greek originated in a particular form of the spoken
language. Now, one of the salient features of the poems is
their wealth of alternative forms with a different metrical
value, e.g. ufaupes and TeaotpES, ‘four*. Doublets of this
type cannot have been a stable feature of any spoken dialect.
They point to an artificial combination of different dialects.
Starting from this point of view, Fick and Bechtel argued that
the poems were composed first in AEolic and then transmitted
to Ionia, where the AEolic forms were Ionicised so far as the
metre permitted . 59 They succeeded in showing that some of the
Aiolicisms are very ancient, and that many forms peculiar
to epic are really AEolicisms in an Ionic dress; but their
attempt to translate the poems back into AEolicwas a failure.
They were left with a number of Ionicisms protected by the
metre, which they could only dismiss as interpolations. The
Ionic element proved ineradicable!
If we admit the possibility that the epic dialect was from the
beginning an artificial medium , 60 we are under no obligation to
confine our search for its origins to the Asiatic coast. In fact,
the archaeological data reviewed in this chapter positively invite
us to the mainland. From this point of view the Arcado-
Cyprian element promises to be specially illuminating, but
first let us dispose of the Attic.
Atticisms protected by the metre are few in number and
67 Nilsson HM 168. 66 Meyer FAG i. 132, GA 2. 75.
60 Fick HO, IUS. «« Meistcr HK.
XVI ARCHEOLOGY AND LINGUISTICS 517
rare.® 1 The accepted view is that they were introduced by Attic
minstrels after the institution of epic recitals at Athens in
the sixth century. 62 Even this may concede too much. Not one
of the rhapsodes mentioned in Attic literature is an Athenian.
They may equally well have been introduced by Ionian min-
strels, whose travels must have given them an acquaintance
with the Attic dialect long before their recitals received
official recognition at Athens. If so, one of the principal argu-
ments for a sixth-century stratum in the poems is invalid.
Aiolic and Arcado-Cyprian are so closely akin that some
authorities treat them as subdivisions of a single dialect,
which they call Achsean. In prehistoric times their relationship
must have been even closer. For the Mycenean age we have to
imagine a form of Greek, divided into northern and southern
subdialects, which extended down the whole coast from Thessaly
to Laconia, excepting Attica, and overseas to Crete, Rhodes,
and Cyprus. This is Nilsson’s view. 63
The following elements in the epic dialect have been clas-
sified as iEolic: the genitive singular in -ao, the genitive plural
in -dcov, the dative plural in -am, the case-ending -91, the
substantives in --rf|p, the infinitives in -pev and -pevai, the
aorist in -crcra, the perfect participle in -ovtes, the pronouns
dppes and Oppes, the apocoptic forms of the prepositions av,
Korr, Trap, the patronymic in -105, the adjectives in -evvos, the
adverbs in -v/8is, the prefix ipi-, the particle p&v, and the
following words: ito-d, ks, Qei, irtavpes» Tay6s, fjporos. 64
The Arcado-Cyprian elements are: tttoAis, tttoAepos (JEo.
Ion. tt6Ais, ttoAepos), (3oAopai (ASo. |36AAopai, Ion. fiouAopai),
t 6 vu (As o. Ion. t65e), and 'the following words: ccCrrdp, 18£,
8£octo, f|TrOoo, 8»pa, k£Aev6os, fjpccp, Aeuaaco, aval;, dvcoyco,
hyn'ip, K^papos, eAos, ypadco, alaa, oloj, efycoAiy 66
The discrepancy between the two lists leaps to the eye. Why
has Arcado-Cyprian contributed so much to the vocabulary of
epic' and so little to the morphology? The truth is that the
61 Wackemagel SUH. 62 Nilsson HM 162.
63 lb. 176. e 4 lb. 163-7.
« 5 BuckGD 132, Bowra HWA. Several of these words are confined in our
records to Cyprus, but presumably they all go back to the Achaean dialect of
the Peloponnese.
518 STUDIES IN ANCIENT GREEK SOCIETY XVI
AEolic list is inflated. The case-endings in -ao and -occov, die
apocoptic forms of the prepositions, the substantives in -T^p,
and die particles pdv and kg should have been entered in both
lists. They are just as much Arcado-Cyprian as AEolic. The
classification of these elements as AEolic is a legacy from the
time when it was taken for granted that the epic dialect
originated in Asia Minor. They are not AEolic but Adhaean.
In regard to the relation between AEolic and Arcado-
Cyprian Nilsson’s argument is straightforward and convincing,
and the adjustment we have just made has strengthened it. It
is only when he turns to Ionic that it becomes blurred:
After the Dorians had broken up the Achasans and driven them out of the
coastal provinces of the Peloponnese, the epic poetry was preserved by the
northern branch of the Achsans, the Aiolians. . . . Finally, when emigrating
to Asia Minor, these brought their epics with them and transmitted them
to the Ionians. 66
If the roots of epic are to be sought on the mainland, then,
as Nilsson himself insists, we must look to the Peloponnese
and above all to Argolis, the centre of Mycenean power. But
the Ionians came from the Peloponnese. Why did their Muse
desert them? Songs are not heavy luggage, even for refugees.
Why did they have to recover them from the AEolians?
This difficulty arises solely from his tacit assumption that
the Ionic dialect had existed as such on the mainland from an
indefinitely remote past. If, as he has shown so lucidly,
AEolic and Arcado-Cyprian had changed during the age of
migrations, the same must be true of Ionic.
The following epic forms are common to Attic-Ionic and
Arcado-Cyprian: (i) si (AEo. ai);'(z) ccv (AEo. kg); (3) the in-
finitive in -vat; (4) T^oaspcs Tour’ (Att. Tf-nrapGs). These
show that Ionic is closer to Arcado-Cyprian than to AEolic.
But are they exhaustive?
The most distinctive feature of the Attic-Ionic dialect is the
vowel shift from a to rj. In Ionic original a disappeared alto-
gether; in Attic it survived only after g, 1 and p. Epic agrees
with Ionic, but with a number of exceptions which have not
been explained. How old was this vowel shift? An answer to this
question may throw light on the linguistic history of the poems.
go Nilsson HM 177.
XVI
ARCHEOLOGY AND LINGUISTICS
519
These epic survivals of 5 have usually been regarded as
.Eolicisms which, though in many cases unprotected by the
metre,, were for some reason never Ionicised. But for what
reason? Why, for instance, does epic always give Aaos, when
Ion. Arj6s would have suited the metre just as well? The question
becomes all the more puzzling when we find vrps ‘temple’ in-
variably in place of vaoj. Why was the procedure so inconsistent?
Let us begin by classifying the examples. They fall into three
categories. In the first we have a forms which are protected by
the metre: the particle pdv (Ion. pev), the substantive Oecc
(Ton. 0eos), and the following proper names: Afvefas, Auyefas,
‘Epuefas, NaUCTiK&a (Ion. *Alv£qs, *Auy£qs» 'EppeqSj
*NcxuCTiKaiT)). Of these the last is specifically AEolic, cf.
’AOavoca for ’AOavata. Secondly, we have the place names
Aapicrcx, Occpos, Oetd (Ion. Aripicra, Ofpos, *®eit|), which may
be referred to the general tendency of place names to resist
dialect modification.
All these instances may be regarded as special cases. Apart
from them, original a survives only before the back vowel
o(co). The answer to our question must be sought in the dialect
history of this double vowel.
In Arcado-Cyprian and ASolic ao is preserved in the middle of
a word. Final -ao becomes -av in Arcado-Cyprian and -a in
iEolic; -aco becomes -a in both. Examples:
Original Arcado-Cyprian JEolic
Aaos
vccuTao
TTocrsiSacov
voutAcov
Aaos
vaurau
rToooiSdv
vauTccv
Aaos
vauTa
TToaefSav
vccOtccv
In Ionic the change was more drastic. The epic forms will be
considered separately. For the moment we are concerned with
Ionic as we know it from other sources. First, do (aco) became
Tp (t|co); then rjo (qco) became sco; and finally the disyllabic eco
was reduced to a diphthong or after a vowel to simple go. This
gives us five stages: (1) *ao (aco); (2) tp (nco); (3) ego; (4) eco;
(5) co. The first stage has disappeared completely. So in many
words has the second. In others the second and third coexist.
Examples:
520 STUDIES IN ANCIENT GREEK SOCIETY XVI
V
I
II
IH-IV
*Aaos
Ar|6s
AecoS
*vaoj
vt)6s
VECOS
*voarrao
vcarrr|o
VCCUTECO
*vauTaoov
*VaUTT|COV
VOCUT^COV
*TToCT£i5acov
*rioCTEtSqcov
TTooeiSegov
*7raiacov
*uaif)cov
*TtaiEGOV
TTOCICOV
The forms marked with an asterisk do not occur. The case-
ending -tjo is found only in early inscriptions. 67 The case-
ending -f|cov is not found at all. Evidently the transition from
the first stage to the third must have been very rapid, especially
in the case-endings.
The treatment of aco (aco) in epic maybe analysed as follows.
In the first category, only the a forms occur. Examples: tAaoj,
dircccov, ’AAkucccov, J A|iu06oov, ’IAoves, Maydcov, TApacov.
In the second category the a and s forms coexist, but without
the intermediate ti form. Examples: Aaos and TTnv£-Aecos,
TTooeiSacov and TTooEiSfcov, genitive singular -ao and -£co,
genitive plural -Acov and -4cov. In the genitive singular the ego
is always monosyllabic; in the genitive plural it is usually so.
In the third category the T| and e forms coexist: vrps and vecoj
‘of a ship’. In the fourth category the r| form is used exclusively,
but this category is very small, being confined to three instances:
vr|6s ‘temple’, Tran) ova, and the proper name Euvrjos. This is a
further indication of the instability of the intermediate stage.
We have already remarked that the classification of these
forms in ao (aco) as JEolic is misleading. They do not occur in
extant AEolic any more than in Ionic. If they are AEolic, they
must be assigned to a prehistoric phase of that dialect, which
we may call proto- AEolic. But they may equally well be
assigned to a prehistoric phase of Arcado-Cyprian. And there
is a third possibility. They may be proto-Ionic. In other words,
they belong to the grec commntt which lies behind the dialect
variants. 68
It seems that the solution of the problem is to be sought
87 GDI. 5423. Reichclt (68) gives the following figures for -ao -eco in the
Iliad and Odyssey: -ao 247 times, -eco before a vowel (where it is replaceable
by -ao) 49, before a consonant (where it is not) 2 7.
88 Mcillct AHLG 163.
XVI ARCHAEOLOGY AND LINGUISTICS 521
along the following lines. The forms tjo (tjco) were transitory.
Before ti had superseded a, it was itself superseded by e. In
other words, original 5, when followed by the back vowel o (co),
had been exceptionally persistent, with the result that, when
it did change, it changed very rapidly. It follows that the sub-
stitution of t) for a was not completed till after the Ionian
migration.
If this conclusion is correct, it enables us to explain why
epic gives vips for va6s but not Atjos for A ao$. The former pre-
supposes an early Ionic *vr|F°S from original *vocfos (cf.
Lesbian vocuos). A digamma has been lost. But Acc6$ (etymo-
logy unknown) shows no trace of a digamma. In this case, being
protected by the following vowel, the a was more persistent.
Before accepting this solution, let us consider whether there
is any other evidence which may help us to date the shift from
« t of.
The Greeks were known to the Assyrians, Persians and other
eastern peoples as Tomans' — the children of Javan, as they are
called in the Old Testament. The oriental forms of the name
all point to a Greek form * 5 Iocfoves. The first recorded con-
tact between Assyrians and Greeks belongs to the year 698 B.C.,
when Sennacherib quelled a revolt in Cilicia. 69 There may
have been others earlier than that, but in any case it seems
that for some time after the migration the inhabitants of
Ionia retained the original 6c in their national name.
The letter H, which, as we know it, denotes long e in con-
tradistinction to E (short e), had originally denoted the
spiritus asper (initial h). The transference was rendered possible
by the fact that the spiritus asper was lost in East Ionic. Now, in
some of the earliest Ionic inscriptions, which date from the
seventh century, the letter H is not used simply for e, as it was
in later times, but only for this derivative e, representing an
original a. Original e is still represented in these inscriptions
by the letter E, which serves also for short e, e.g. NIKANAPH
NT ANEGEKEN EKHBOAOI IOXEA 1 PHI.™ This distinc-
tion must mean that derivative e had not yet become identical
with original e. In other words, the shift from a to e was not
69 Pauly-Wissowa s.v. Iones; King SI, Cuny 21.
79 Buck CGGL 72, Lejeune 205.
522 STUDIES IN ANCIENT GREEK SOCIETY XVI
yet complete. And that being so, there is no difficulty in sup-
posing that original a had persisted in certain positions a
couple of centuries earlier.
On the other hand, the shift must have begun before the
Ionian migration. That is clear from its presence in Attic. But
in Attic its range is restricted. It does not operate after s, i or p.
We may infer that its extension beyond these limits took
place in Ionia after the migration. This too is in keeping with
the supposition that Ep. ao (aco) is early Ionic.
Finally, if the shift had been much older than the migration,
we might expect to find some vestiges of it on the mainland,
where the emigrants came from. The nearest approach to it is
in the Boeotian dialect, which gives regularly t] for on, but not
Ti for a. In the Peloponnese we find several correspondences
between Ionic and Arcado-Cyprian, but not this; we find
several Arcado-Cyprian survivals in Argive and Laconian
Doric; we find forms which, though not specifically Arcado-
Cyprian, are recognisably ‘Achaean’ in the North-West
dialect of Elis; but nowhere in the Peloponnese or in any
other part of the mainland except Attica do we find .the
slightest trace of T) for original a. If we recall what Herodotus
says about the circumstances of the migration — that the
colonists were fugitives from many different parts of Central
Greece and the Peloponnese, speaking a variety of dialects, and
that their point of departure was Attica — we have a strong
case for regarding the beginning of this vowel shift as an in-
cident in the upheaval created by the Dorian invasion . 71
I have dealt with this problem in some detail, because it
throws light on a further problem: what was the relationship
between the parents of Attic-Ionic and Arcado-Cyprian before
the migration? One of the elements common to Attic-Ionic
and Arcado-Cyprian is the particle ccv (JEo. kg). Here
Arcadian differs from Cyprian, which has only kg, like AEolic.
Arcadian has regularly ccv, but it uses ke in the phrase ei k’ ccv,
which is a combination of ai ke and si ccv, designed to
obviate the hiatus . 72 This indicates that av was intrusive in
Arcadian — a borrowing from Attic-Ionic; and in that case we
must suppose that an early form of the latter — proto-ionic —
71 Cf. Lcjcunc 17. 72 Buck GD 98.
XVI ARCHAEOLOGY AND LINGUISTICS 523
was spoken alongside of Achaean in the pre-Dorian Peloponnese.
I suggest that this is the dialect referred to in the ancient
tradition that Ionic had once been spoken in the north and
north-east of the Peloponnese; 73 and that it was introduced
by the Lapithai, whose presence we have traced in this very
region (p. 264). There it was overlaid by Achaean, which was
contaminated with it.
Table XVU
This reconstruction enables us to account for another
peculiarity of the epic dialect. For the infinitive of athematic
verbs it has three alternative forms with different metrical
values: -nev, -vcci, -uevai. The first is mainland /Eolic
(also Doric and North-West Greek); the second is Arcado-
Cyprian and Attic-Ionic; the third is confined to Asiatic
73 Hdt. x. 145-6, 7. 94 (north coast), 8. 73. 3 (Kynouria), Paus. 2. 26.
x (Epidauros), Str. 392 (Megara).
524 STUDIES IN ANCIENT GREEK SOCIETY XVI
AEolic. I infer that -jjisv was the original Achaean . form, that
-vai was borrowed by the Peloponnesian Achaeans from proto-
ionic, and that -usvou was a conflation of the two. An exact
analogy is available in the Rhodian infinitive in -new, which
is a combination of -piev with the thematic infinitive in -eiv. 74
In prehistoric times, then, an Achaean dialect was spoken in
Thessaly, Bceotia, the Peloponnese, Crete, Rhodes, and
Cyprus. It was divided into two branches, northern and
southern, the parents of JEolic and Arcado-Cyprian. The
southern branch was affected by the speech of the Lapithai,
already established in Attica and the northern Peloponnese.
At tire end of the second millennium West Greek dialects,
closely related to ASolic, were introduced by die Thessaloi,
Aitoloi, and Dorians. Meanwhile Achaean had been carried
across the AEgean to Anatolia, where the two branches emerged
as Asiatic AEolic and Ionic. The former was free from West
74 Buck CGGL 305.
XVI ARCHAEOLOGY AND LINGUISTICS 525
Greek but became contaminated with Ionic, especially in the
middle region of Smyrna and Chios. Ionic was from the be-
ginning a mixed dialect, formed by the fusion of Achaean,
which had already been subject to proto-ionic influence, with
a variety of local vernaculars, including Attic.
For die purpose, of fixing the chronology of the poems the
linguistic approach is complementary to the archaeological. It
is die only other test that can be applied with any confidence
in the results. Arguments from style, plot, characterisation,
and other literary considerations are from the nature of the case
unverifiable. They are drawn from our own preconceptions of
what an epic should be, and, since in our literature the art of
epic, even written epic, has been dead for centuries, they are
inherently unreliable. This is not to say that they must be
excluded altogether,: merely that they must be deferred until,
. having analysed the concrete data, we have discovered what
Homeric poetry was and where it differs from our own.
Linguistics have the same concrete bearing on the form of the
poems as archaeology on their content. But the evidence they
provide is purely relative. There are no documents, con-
temporary or older, with which the language of the poems can
be compared. It can only be compared with itself. We can
enquire whether the Iliad is older than the Odyssey , whether
some books are older than others, and how they stand in rela-
tion to Hesiod and the Homeric Hymns; but the most that
' can emerge is a sequence, which must be fixed by other means.
A great deal of work has been done along these lines. The
relative frequency of various forms and usages in the two poems
and in different parts of each has been studied repeatedly with
a view to determining their date. At the beginning of the
present century the results were generally held to confirm the
-separatist position, but since then Stawell and Scott have
drawn attention to some serious errors in the calculations of
their predecessors and produced new estimates of their own
which point in their opinion to unity of authorship .* 6 Setting
aside the question of authorship, their results tend to suggest
that the Odyssey is on the whole a little later than the Iliad but
75 Scott UH 83-105, RAIO, RAHB, Shewan LD, Stawell HO
• 93-t°4-
526 STUDIES IN ANCIENT GREEK SOCIETY XVI
much earlier than the Hesiodic poems and the Homeric
Hymns. This accords with the result of Wackemagel’s study
of the Atticisms. There are 18 Atticisms in the Iliad (1 in 871
verses) and 26 in the Odyssey (1 in 637 verses). 76
I believe that this conclusion is correct. If I hesitate to
accept it without reserve, it is because neither Stawell nor
Scott was primarily a specialist in the Greek language, and
their results are incomplete. What we need is a comprehensive
study of the whole field of Homeric Greek at the level of
Wackernagel's Sprachlichen Untersuchmgen zu Homer, carried out
by scholars who approach the problem, like Wackernagel,
without prejudice to the question of authorship.
Homeric linguistics are thus, so far as they go, in harmony
with Homeric archeology. The nucleus of the epic tradition
was a heritage from the Mycenean age. Transplanted to
Asia, it was worked up, probably in the neighbourhood of
Smyrna and Chios, by poets whose ancestors had come partly
from Thessaly and Boeotia, partly from the Peloponnese. Their
spoken language, thrown into confusion by the migrations,
was exceptionally rich in parallel forms, and they seized on
this natural advantage to elaborate an eclectic medium
transcending dialect boundaries and distinguished by its
metrical fluency. It was stabilised eventually on the basis of
East Ionic, which in Smyrna and Chios prevailed over AEolic.
The surviving non-ionic forms were not felt as such, being
integral features of the medium that had grown out of them,
and were still freely used in expanding the poems. But, where
metrical considerations did not intervene, the language
assumed an Ionic colouring, and, as the minstrels travelled
further afield, they admitted new forms from West Ionic and
Attic, in accordance with the catholic principle inherent in
their tradition. And so in this unique set of linguistic condi-
tions they brought to perfection a superb vehicle for narrative
poetry, embodying a felicitous union between nature and art.
70 Wackernagel SUH. Their distribution is as follows (instances marked
by him as doubtful are counted as £): ll. II 4, HI i£, IV £, V £, VII x,
Xh XI 3, xn b XIH h XIV b XV i\, XIX 1, XXI h XXIII z\, 01 1 1,
HI 2, IV lb VI b VII lb Vin 4, XI i, XIV 1, XV zb XVI i, XVIII 1,
XVIII 2, XIX lb XX Zb XXI 1, XXIV 2.
XVI
ARCHEOLOGY AND LINGUISTICS
527
5. The Epic Style
Heroic poetry is devoted to remembering the deeds of
famous men. It is individualistic, aristocratic, patriarchal,
martial. Its distinctive qualities are all characteristic of a
definite historical stage — the emergence of the class-struggle.
The particular conditions in which the class-struggle develops
vary in each case, and the most favourable to heroic poetry are
those in which the transition is rapid and abrupt — when back-
ward peoples, after nurturing social inequalities within the
tribal system through contact with a superior culture, are
driven by these internal stresses to plunder and conquer their
civilised neighbours, appropriating their riches and their art.
. Such in brief was the history of the Germanic tribes that
pressed against the Roman frontiers. When we first meet
them, in Caesar's Commentaries, they are still tribal; in the
pages of Tacitus they have perceptibly advanced; and a few
generations later they are carving kingdoms out of the imperial
provinces. We know from Tacitus that they cultivated ancient
songs, in which the memory of great leaders like Arminius
was kept alive. 77
In his account of early TEgean piracy Thucydides remarks
with his usual insight that the motive behind the raids was the
thirst of the leaders for personal gain combined with the need
to provide for their poorer followers. 78 These brigand chiefs
were the founders of the Mycenean monarchies, and among
the by-products of their career was the art of epic. The
rapidity of their rise to power explains why Greek epic differs
so sharply from the choral dances out of which it had evolved.
In art, as in life, there was a violent break with the past. But
for the same reason Greek epic preserves, in common with
Beowulf and the Eddas, many primitive characteristics that have
disappeared in the so-called ‘literary' epics of mature class
society. The contrast is all the more striking because- Homer
has dominated the European tradition. Virgil modelled him-
self on Homer, Dante on Virgil, Milton on both; but there are
certain features of the Homeric style which they never
77 Tac. G. z, Ann. 2. 88. So among the Gauls: Amm. Mate. 15. 9. 8.
78 Th. x. 5. x.
528 STUDIES IN ANCIENT GREEK SOCIETY XVI
attempted to reproduce. This is because in one essential
respect it was alien to them and inimitable. It was preliterate.
The problem of writing was the starting-point of the
Homeric controversy. In the eighteenth century it was argued
by Vico and others that, at the time when die poems were
believed to have been composed, the art of writing was un-
known. This idea was taken up by Wolf in 1795, when
' Europe was in the ferment of the French Revolution. 79 While
yielding to none in admiration of the poems, Wolf argued that
they were compilations of shorter lays put together at Athens
in the sixth century. His views were developed by Lachmann,
who anatomised the Iliad, and a litde later Kirchhoff applied
the same method to the Odyssey . 80 As the nineteenth century
wore on, the Homeric Question broke all bounds and became
a happy hunting ground for graduates in quest of a doctorate.
The ne plus ultra was reached when Wilamowitz discovered
that the Iliad was f a miserable patchwork’ and Fick condemned
the Odyssey as ’a crime against the human intelligence’. 81 The
opposition was cowed into silence. In course of time/ however,
it became apparent that, so foe from reaching any unanimity
among themselves, the separatists had only succeeded, between
them, in condemning the whole of both poems as an inter-
. polation. In the present century, taking courage from this un-
comfortable result, the Unitarians have counter-attacked and
proclaimed complete unity of authorship as boldly as the
separatists have denied it:
It is possible to believe that Greece had one man who could project such
mighty, such enormous works of art, but it is unthinkable that she had at
any period two men, or a group of men, with any such capacity. 82
Students of bourgeois thought will recognise the fallacy in-
herent in this arid controversy. One school of bourgeois
historians has sought to explain all human progress in terms of
the conscious activities of individuals; another' has reduced it
to the operation of inexorable economic forces. 83 So the
79 For the history of the Homeric controversy see Jebb 103-55, Nilsson
HM 1-51.
so Lachmann BHI, Kirchhoff HO.
81 Wilamowitz IH 322, Fick EO 168.
82 Scott UH 268-9.. 89 Plckhanov RIH.
XVI ARCHAEOLOGY AND LINGUISTICS 529
Unitarians see in Homer simply a supreme example of the
miracle of genius, while for the separatists he is the eponym of
an arbitrary collection of Yolhlitder, which had sprung up
anonymously and artlessly on the lips of the people. It is the
old bourgeois dilemma, the stale antithesis between idealism
and mechanical materialism.
In the last fifty years, thanks to Schliemann and Evans, it
has become known that the art of writing was practised in the
Aegean as far back as 2500 B.c. This was the Minoan script.
The Greek alphabet was introduced by the Phoenicians, per-
haps as early as the eighth century. Meanwhile it has been
placed on record that in the year 1887, between January 2 and
February 15, a Croatian minstrel recited from memory at
Agram a series of lays amounting to twice the combined
length of the Iliad and O dysstyM Thus, whether early or late,
Homer may have been literate, and, even if he was not, he
might still have composed the poems that go under his name.
Before reaching a decision on this point we must examine his
work in the light of other traditions of heroic verse which
are known to have been transmitted entirely by word of
mouth.
The Kirghiz are to-day free and equal citizens of the Kirghiz
Republic, which lies in the Tien Shan Mountains north of the
Hindu Kush. Before the Revolution of 1917 they were back-
ward, disease-ridden nomads, doomed apparently to extinction,
but famous for their poetry. They are still famous for their
poetry, though in all other respects they have been transformed.
The following account is from nineteenth-centuty travellers
who knew them in their primitive state. 86
They were all poets. Almost everyone was able to improvise
heroic verse, though only professionals performed in public.
These travelled the country, reciting at festivals and accompany-
ing themselves on a two-stringed instrument called the koboj?.
Every local khan had his own minstrel, whose task was to
celebrate his achievements.
One of these minstrels attached himself to a Russian ex-
peditionary force sent into Kirghizia in 1 860:
84 Murko 284.
86 For the quotations that follow I am indebted to Chadwick GL 5. 174-91.
53 ° STUDIES IN ANCIENT GREEK SOCIETY XVI
Every evening he attracted crowds of gaping admirers, who listened avidly
to his stories and songs. His imagination was remarkably fertile in creating
feats for his hero — the son of some khan — and took daring flights into the
realm of marvel. The greater part of these rapturous recitals he improvised,
the subject alone being usually taken from tradition. 88
Their technique is described by Radlov:
One sees from a Kirghiz reciter that he loves to speak, and tries to impress
his hearers by elaborate strophes and well-turned phrases. It is obvious too
on all sides that the listeners take pleasure in them and can judge if an ex-
pression is well rounded off. Deep silence greets the reciter who knows how
to arrest his audience. They sit with head and shoulders bent and shining
eyes. They drink in his words. Every adroit expression, every witty word-
play calls forth lively applause. . . .
Every minstrel with any skill at all always improvises his songs on the
spur of the moment, so that he is not capable of reciting a song twice over
in exactly the same form. But it must not be supposed that this means that
he composes a new poem each time. The procedure of the improvising
minstrel is exactly like that of the pianist. As the pianist puts together in
harmonious form various runs that are familiar to him, with transitions and
motives according to the inspiration of the moment, and thus makes up the
new out of the old, so also does the epic minstrel. Thanks to long practice,
he" has a whole series of ‘elements of production’, if I may so express it,
which he puts together in suitable form according to the course of the nar-
rative. These consist of pictures of certain events and situations, such as the
birth of a hero, his growing up, the glories of weapons, preparations for
fighting, the storm of battle, the conversations of a hero before battle, the
portrayal of people and horses, the characterisation of the well-known
heroes, the praise of the beauty of a bride. . . . His'art consists in piecing
together these static components as circumstances require and connecting
them with lines invented for the occasion. All these formative elements he
can use in very different ways. He knows how to sketch a picture in a few
strokes, or paint it more thoroughly, or elaborate all the details with epic
fulness. The more of these elements he has at his disposal, the greater the
diversity of his performance, and the greater his power to sing on and on
without tiring his listeners with a sense of monotony. A skilled minstrel can
recite any theme he wants, any story that is desired, extempore, provided
only that the course of events is clear to him. When I asked one of their
most accomplished minstrels if he could sing this or that song, he answered:
‘I can sing any song whatever, for God has implanted this gift of song in my
heart. He gives the words on my tongue without my having to seek them. I
have learnt none of my songs. All springs from my inner self.’ And the man
was right. The improvising minstrel sings without reflection, simply from
80 J and R. Michell 290.
XVI ARCHAEOLOGY AND LINGUISTICS 53 1
his inner being, as soon as the incentive comes from without. ... He can
sing for a day, a week, or a month, just as he can talk, and narrate all the
time.
Lastly, just as the gift of poetry was common to the whole
people, so the characteristics of poetical diction were present at
a lower level in their ordinary speech:
The words of every Kirghiz roll tripping off the tongue. Not only has he
sufficient command of language to improvise long poems, but even his
ordinary conversation shows traces of rhythm and artificial arrangement.
His language is figurative, his phrases sharp and clear-cut. 88
Radlov has revealed the secret of the minstrel's art. In verse
words are arranged in artificial patterns, and, if the minstrel
is as fluent in this medium as he is in common speech, it is
because he has at his disposal a repertory of traditional
formulations, covering all the themes incidental to his subject,
all the prescribed rituals and procedures of primitive life. All
this he has acquired along with the rest of his craft. The epic
style is facile precisely because it is formal. Its conventional
character is derived from its origin in improvisation . 89
These features are universal. Just as the social setting of
these Kirghiz minstrels reappears in the palace of Odysseus,
so their use of language is echoed in the Iliad and Odyssey . Or
again, if we compare, as Chadwick has done, Greek epic with
Germanic, we find the same use of static epithets, figurative
tropes, and repeated paragraphs for describing such actions as
going to bed, getting up, preparing meals, receiving strangers,
harnessing horses. As he has remarked, ‘both sets of poems
were designed for preservation by oral tradition . 90
The presence of such features in the Iliad and Odyssey is
proof that the poems had grown out of conditions such as
Radlov has described, but of course this does not prevent us
from believing that, as we have them, they belong to a far
8 ’ Radlov PV 5. iii, xvi. As Radlov points out, the minstrel adapts
his performance to the nature of the occasion. Thus ‘if rich and dis-
tinguished Kirghiz are present, he knows how to introduce panegyrics veiy
skilfully on their families. ... If his listeners are only poor people, he is not
ashamed to. introduce venomous remarks concerning the pretensions of the
rich, and in greater abundance according as he is gaining the assent of his
listeners’ (PV 5. xviii-xix).
88 Radlov AS 1. 507. "Cf. ChadwickGL 3. 669. 90 ChadwickHA 320.
532 STUDIES IN ANCIENT GREEK SOCIETY XVI
higher order. They do undoubtedly belong to a higher order,
but, if we are to understand how they became what they are,
we must tread cautiously, making sure of every step. Is there
anything in Homer's handling of these ‘elements of production’
to suggest that he has advanced beyond this preliterate stage
of a still fluid oral tradition?
As Bowra has remarked, they have a functional value for the
listeners. 91 In fact, they are as necessary for the audience as
they are for the poet. Just as they enable him to compose with
fluency, so they relieve the strain of listening by interposing
words, phrases, paragraphs, which, being familiar, permit
momentary relaxations of attention. They would consequently
have tended to survive for as long as the poems continued to be
publicly recited. This is true, but, since they had the same
function in primitive conditions, it does not help to decide
the point at issue.
A static epithet adds very little to the meaning. That is its
virtue. It follows that, taken literally, it will be less apposite in
some contexts than in others. This did not trouble the primitive
minstrel, but sophisticated poets are more fastidious. It will
therefore be a sign of advance beyond primitive technique if
it can be shown that Homer uses these static epithets dynamic-
ally — with conscious regard for the context.
In the Odyssey the story of Clytemnestra’s adultery is intro-
duced with an allusion to ‘faultless Aigisthos’ (1.29). The
adjective is used elsewhere in a fully active sense. Here it would
be grotesque if it were anything more than static.
In Iliad XVI (298), as he scatters the clouds from the hill-
tops, Zeus is designated 'lightning-gatherer'. The use of this
epithet in preference to the commoner ‘cloud-gatherer' has
been hailed by Bowra as an instance of discrimination. 92 It
does not take us very far. To call Zeus ‘cloud-gatherer' when
he is scattering the clouds would surely have been too much
even for a primitive minstrel. Moreover, the other one, though
not so common, was' also traditional; 93 and, while it avoids a
flat contradiction, it is not really any more appropriate.
The other cases adduced by Bowra need hot be examined in
detail. There is nothing in them. Thus, when Diomedes ‘good at
91 Bowra TDI 81. 22 lb. 83. os cf. II. 1. 580, Hes. Th. 390.
XVI ARCHAEOLOGY AND LINGUISTICS 533
the battle-ay’ stands shuddering at the onset of the war-god,
we are to understand that ‘the epithet, so far from being
superfluous or inappropriate, gives exactly the right idea of a
brave man being for once afraid'. 94 It seems to me that
Diomedes here is on a par with Menelaos, who is ‘good at the
battle-cry' when he is getting out of bed. 95 What Bowra fails
to notice is that, had these epithets been used, even occa-
sionally, as he supposes, they could not have preserved their
function as rests. The ear would have been constantly on the
alert for some jeu d’ esprit. The Homeric treatment of this
feature is in strict conformity with primitive usage.
After Homer there was a change. In the Catalogue of Women,
in a list of Helen’s suitors, we meet the line (29):
hi 6’ ’Ifl&KT); luvorro 'OSuo-crfjos lepfj Is.
In Homer lepfj Ts is a static epithet of Telemachos, never of
Odysseus. But here Hesiod is thinking of Odysseus as a young
man, and accordingly he feels that the Homeric ttoAutAcxs
6ios ’OSuctcteus, which would anticipate his future, will not do.
So he describes the father as Homer had the son. This is really
original, but it belongs to the decadence of the epic tradition.
We notice that in adapting the formula iepfi Ts to Odysseus
Hesiod has admitted a false quantity. He treats the final
syllable of ’OSuacrfjos as long. This brings us to our second
test. Besides defying the sense, these formulas often conflict
with the metre.
The poems are full of metrical anomalies. Many can be
explained by the loss of the digamma, which went out of the
spoken language while they were taking shape. But the
Homeric treatment of the digamma is inconsistent. Even in
the same word it is sometimes functional, sometimes not:
1 L 2.373 TTpt&poio ocvccktos, 24.449 TTolqaccv avoncm.
The explanation is that, after becoming obsolete, it was
treated in some cases as justifying the resultant false quantity
or hiatus, while in others, quick as always to utilise alternative
metrical values, the minstrels followed the spoken usage and
ignored it.
There are, however, a great many anomalies that cannot be
94 Bowra TDI 84. 95 Oi. 4. 307.
534 STUDIES IN ANCIENT GREEK SOCIETY XVI
referred directly to this source. Thes^e are due to analogy. Once
established in the special conditions arising from the, loss of
the digamma, the principle of false quantity or hiatus was
extended. This point has been elucidated by Milman Parry,
who has shown from hundreds of examples how it enabled the
minstrels to enhance the adaptability of their 'elements of
production'. 96 Take the following verses:
Od. 2.2: <6pvirr’ dp’ £§ euvfj^iv ’OSi/aafjos 9IX0S utdj.
Od. 3.305: dSpvuT* dp’ §§ £uvfjq>i repfivioj lirrrdra NSoTCop.
Od. 4.307: &pwx’ dp* s0vfj9i pof)V dyaSis Mev&aoj.
Od. 15.59: *riv 8’ d>s oOv £v6t\cev 'OSuaafios 9O.OS utds.
IZ. 3.21: t6v 8’ cos o 5 v £v6qaEV dpt^iXo; Mev&oos.
II. 5.95: t6v 8’ d>s odv tvdrjCTE Auxdovos iyAai; u! 6 s.
II. 21.49: t6v 8’ cbs ov 5 v Jv>6r]ff£v 'AxtMfja -nroMiropSov.
II. 21.415: t6v 8 s &s ovjv £v6t]oe 0e& ^EUKcbXEVos *Hpt).
Od. 19.59: tv0o KctO^ET 1 firerra mptypcov 17 rivEX 6 Treiot.
Od. 19.102: tvOct kcx 0 ^et’ lireiTa noAdrAots Bios ’OSuctoeus.
These verses, all composed entirely of formulas, are regular.
But we also find:
Od. 16.48: n> 0 a koQ^et’ tireiTa * 05 u<xafjoj 9(Xos v!6j.
Here we have an irrational hiatus. Being familiar from constant
usage in their respective positions in the verse, the two
formulas are juxtaposed in defiance of the metre. Irregularities
of this type abound, being both natural and necessary if the
poets were to have a free hand with their 'elements of pro-
duction*.
‘For Homer’, Party wrote, ‘as for all minstrels, to versify
was to remember — to remember words, expressions, phrases
from the recitals of minstrels who had bequeathed to him the
traditional style of heroic verse’. 07 How true this is may be
judged by examining continuous passages. Of the first 50 lines
of the Iliad no less than 36 are constructed wholly or partly
out of phrases which can be recognised as ‘elements of pro-
duction.’
Parry denied to Homer all originality of style. This has
shocked the literary critics, but in the sense in which he
intended it it is quite correct. Homeric diction is traditional,
96 Parry FMH, SET. e? Pany FMH 6.
XVI ARCHAEOLOGY AND LINGUISTICS 535
not individual. It conforms to the conditions of oral recitation.
But style is an elusive thing. Old materials may be put
together in new ways. Old conventions may be qualitatively
refined without any overt departure from accepted methods.
We have seen that the epic dialect, produced by a specific set
of objective linguistic conditions, was organised and expanded
by the poets with a conscious realisation of its intrinsic
potentialities. The same is true of the epic style.
One of the features that distinguishes the Iliad from other
early epics is the copious use of similes. The simile is of course
used in Germanic epic and in the same way, but on a much
smaller scale. In the Iliad it is highly organised and worked
into the structure of the poem . 08
The majority of Homeric similes are taken from country
life. They present a consistent picture of a simple, sedentary
society dependent on pastoral husbandry. It has been sug-
gested that they are relatively late, referring to the poets’ own
time rather than the heroic past . 00 Their general characteristics
are well known. They tend to repeat themselves, often word
for word; they are often elaborated beyond the point of con-
tact with the reality; and some of them are frankly irrelevant.
In this they resemble the static epithet. Just as the epithet
relaxes the attention, so the simile provides a diversion. It is, at
least in origin, an ’element of production'.
In Book II the Achaeans are preparing for battle:
As a fire rages through the woods on a hilltop, visible from afar, so the
gleam of bronze flashed to heaven. As when many breeds of fowl, geese or
cranes or long-necked swans, fly to and fro over the water-meadows of
Kaystros and settle screaming in an uproar, so the Achaeans poured from
their tents and ships into the plain of Skamandros, and the ground clattered
under the feet of men and horses. They stood in thousands in the river
pastures like the leaves or flowers of spring. mo
So far all is straightforward, but then:
Like flies in spring that hover thick in the farmyard when the milk spurts
into the pails, so many were the Achaeans as they took their stand in the
plain to meet the Trojans. 101
This adds nothing, rather detracts. The flies are still hovering
when the troops take up their positions. Moreover, they hover
08 Sheppard SI. °°FrankelHG. 100 Z/. 2.459-68. 101 ll. 2.469-73.
536 STUDIES IN ANCIENT GREEK SOCIETY XVI
again in Book XVI, where the same simile is repeated . 102
Intent on enhancing the impressiveness of the occasion, the
poet has drawn on his repertory a little too freely forour taste.
The same thing seems to have happened in the famous
passage where Hector is running from Achilles:
It was a brave man that ran, chased by a braver. And not for the prize in a
footrace — a beast for sacrifice or an oxhide shield; they were running for
Hector's life. As champion horses round the post at full gallop, with a prize
from a dead man's treasures, a tripod or a woman, awaiting the winner
at the goal, so these two sped round Priam's city. 103
FIG. 82. Footrace: Attic vase
Again the simile is traditional; we have heard it already in the
same book . 104 And there is little point in comparing them to
racehorses after the closer parallel of the footrace.
For fear of misunderstanding let me explain that I am not
dismissing these less exact similes as interpolations. If, as I
believe, they are 'elements of production’ drawn from a com-
mon store, they belong to a time when the poems were still
fluid — when they had never been recited twice over in pre-
cisely the same form. They belong to a phase in the evolution
of poetry which by its very nature excludes the possibility of
interpolation — or, what comes to the same thing, to a phase in
ioa U. 16. 641-3. i °3 II. 22. 158-66. i° 4 11 . 22. 22-3.
XVI ARCHAEOLOGY AND LINGUISTICS 537
which all poetry was nothing but interpolation. And in those
conditions they were justified. If these minstrels had not learnt
the lesson that later Greek poets had to learn — ‘to sow with the
hand and not with the sack’ 106 — the reason is that they were
working in a different milieu. The economy of detail recommended
in this maxim would have overtaxed the listener's attention.
So far, then, the Homeric simile seems to mark no advance
on primitive technique. But it did not stop there. Besides these
artless similes, used lavishly for broad effects, we find others so
vivid in their accuracy that they have been the envy of poets
ever since. Every reader will have his own favourites. Mine comes
from the passage already cited, when Hector can run no further:
As when a hound makes after a kid in the mountains after starting it from
its lair and chasing it through glen and valley, until it crouches under a
thicket in the hope of escape, but sticking to the trail the hound noses it out,
so Hector tried in vain to elude Achilles.
This, though quite effective, is traditional. We have followed
such chases many times. But the poet has not finished yet:
As in a dream the pursuer cannot overtake nor the fugitive escape, so
Achilles could not catch Hector nor Hector get away.ioc
This is perfect. It does not divert, but illuminates the object.
And there is nothing else like it in the Iliad. It has the air of
conscious art.
We get the same impression even more strongly if we take a
series of related images. One well-defined category is de-
signed to illustrate the descent of deities from Ida or Olympus:
Thus she spoke, and Athena was stirred to act. Down she darted from the
peaks of Olympus like a star sent by the Son of Kronos as a sign to sailors
or fighting-men, a bright shooting star that trails a shower of sparks. So
Pallas Athena darted down to earth. 107
Thus she spoke, and wind-footed Iris obeyed. She descended the peaks of
Ida down to Troy, as when a chill srorm of snow or hail sweeps from the
clouds under a northerly blast. So swiftly did Iris fly.108
The formula of introduction and conclusion is almost the same
in each case, and the similes themselves might be interchanged
without damage to the context. Tradition demanded a simile
at these points. A primitive poet would have been content to
10s Piu. c/or. Ath. 4. 106 II. zz. 189-201.
107 II. 4. 73-8. 108 II. 15. 168-72.
538 STUDIES IN ANCIENT GREEK SOCIETY XVI
repeat the same one, just as he repeats die static epithet.
Homer preserves the form but varies the content. He uses
V
fig. 83, Iris: Attic vase
the convention as a pretext for inserting a vivid diversionary
vignette. And sometimes he is more venturesome:
Thus he spoke, and white-armed Hera obeyed. From the hills of Ida she
darted to Olympus, and, as a man who has travelled far turns over many
thoughts in his mind, musing rapidly, ‘If only I were there or there!’ so
rapidly did Hera fly. 100
This is a witty elaboration of the formula ‘as swift as a wing
or a thought *. 110 But beyond the idea of speed it has nothing
to do with Hera, who is not at present in a reflective mood —
100 II. 15. 78-83. no ol 7. 36.
XVI ARCHEOLOGY AND LINGUISTICS 539
she is in a very bad temper. It is as though, tired of formalities,
the poet has decided boldly for an imaginative novelty.
These instances are still perfunctory in the sense that they
are merely designed to illustrate the habitual movement of die
immortals. If in effect they are more than that, they owe it to
the poet’s originality. And when the occasion calls for some-
thing out of the common, he rises to it. The story of the Iliad
turns on the disastrous quarrel between Agamemnon and
Achilles, which began with the plague sent by Apollo:
Thus the priest spoke in prayer, and Phccbus Apollo heard him. In fury
he descended from the peaks of Olympus, his bow and quiver on his
shoulders, his arrows whistling angrily as he moved. He came like the night. 111
The formal introduction has disappeared, and the simile has
been reduced to an afterthought. That makes it all the more
impressive. This is mature art — a fine example of die free
handling of an inherited convention.
Nearly all the similes in the Iliad occur in the battle scenes,
where they lend colour and variety to the grim catalogues of
slaughter. This was certainly deliberate. Not only are the
interludes, such as the deception of Zeus and the embassy to
Achilles, almost free of them, but in the Odyssey , which has a
more varied and homely plot, there are hardly any similes at
all. Here then is a real instance of artistic discrimination,
testifying to a sustained sense of unity. The Homeridai did
more than transmit. In transmitting they transformed. They
were all hereditary craftsmen, but the best of them were
creative artists. Yet even these exercised their originality in
refining and harmonising their technique rather than in radical
innovations. The Iliad and Odyssey are made of the same stuff as
primitive epics, and made in the same way, but in them the
qualities inherent in improvised verse have been nursed up to
the point at which, without losing any of their spontaneity,
they blossom into art. The easy effortless mastery that makes
the Homeric style so brilliant was the result of many centuries
of practice, cultivation, and refinement.
All aesthetic judgments turn ultimately on personal experi-
ence, so let me explain how some of my own misunderstand-
ings of Homer came to be cleared up.
111 1/. 1. 43-7.
540 STUDIES IN ANCIENT GREEK SOCIETY XVI
I read the Odyssey first, and like every schoolboy I was thrilled
when I came to the untranslatable lines:
ail 8*£v CTpo<p6?uyyi kovIti;
KElao pfya; peyaXcooTl, XiAaapfvo; linrocjuv&cov. 112
This was magnificent, inspired. In due course I came across the
same line in the Iliad. This was most perplexing. If it was really
inspired, how did it bear repetition? It was no comfort to be
told that one passage was an imitation of the other, because
then the poems were indeed a patchwork in which it was im-
possible to distinguish the counterfeit from the real thing.
After encountering other repetitions of the same kind, I put
them all down as ‘primitive’, but without understanding what
that meant.
Then I went to Ireland. The conversation of those ragged
peasants, as soon as I learnt to follow it, electrified me. It was
as though Homer had come alive. Its vitality was inexhaustible,
yet it was rhythmical, alliterative, formal, artificial, always on
the point of bursting into poetry. There is no need to describe -
it further, because it had all the qualities noted by Radlov in
the conversation of the Kirghiz. One day it was announced
that a woman in the village had given birth to a child. As my
informant expressed it, Ta si tarraigthe aniar nice , ‘She has
brought her load from the west’. I recognised the allusion,
because often, when turf was scarce, I had seen the women
come down from the hills bent double under packs of heather.
What a fine image, I thought, what eloquence! Before the day
was out, I had heard the -same expression from three or four
different people. It was common property. After many similar
experiences I realised that these gems falling from the lips of
the people, so far from being novelties, were centuries old —
they were what the language was made of; and as I became fluent
in it they began to trip off my own tongue. Returning to Homer,
I read him in a new light. He was a people’s poet — aristocratic,
no doubt, but living in an age in which class inequalities had not
yet created a cultural cleavage between hut and castle. His
language was artificial, yet, strange to say, this artificiality was
natural. It was the language of the people raised to a higher
power. No wonder they were enraptured.
“ 2 01. 24. 39-40, II. 16. 775-6.
xvn
THE HOMERIDAI
I. Aiolis and Ionia
The site of Troy, commanding the highways between two
continents and two landlocked seas, had attracted settlers from
very early times. 1 Troy I was an open village with a neolithic
culture indigenous to Anatolia. Troy II was fortified, with a
central palace of the same type as those of Dimini and Sesklo
(p. 185). It was a thriving market town, importing hammer-
axes from central Europe, razor-blades from the Caucasus, and
pottery from the Cyclades. About 2000 B.c. it was razed to the
ground. The date coincides with the rise of the first Hittite
kingdom in Cappadocia, and perhaps with the first appearance
of Pelasgoi in the AEgean basin (pp. 193, 261). Two small
villages (Troy Ht-IV) lingered on amidst the ruins. Then the
city was rebuilt and enlarged (Troy V-VI). Troy VII was the
Homeric city. Minoan influence was now dominant. The Troy
which the Achaeans sacked belonged to die same culture as their
own Mycenae.
After that Troy disappeared. Historical factors combined to
nullify its natural advantages. The recrudescence of piracy had
put a stop to the Hellespondne sea-traffic; the Hittite Empire
had collapsed; and a litde later the trade route across the land-
bridge was cut by the Phrygian invasion of Anatolia. Fuit
Ilium.
The earliest Greek settlers on this side of the AEgean did not
go there in pursuit of trade. They wanted a place to live in.
The most desirable part of the coast was the stretch between
the Hermos and theMaiandros. Here harbours were plentiful
and commodious, with luscious pasturage in the lower valleys,
while higher up caravan routes penetrated into the Anatolian
highlands, conveying merchandise to and from Cappadocia,
Syria, and the infinite empires of the east. Here Minoan and
1 Childe DEC 3 5.
542- STUDIES IN ANCIENT GREEK SOCIETY XVII
Hittite influences had converged to produce among the
Cafians and Leleges a culture less brilliant than the Mycenean
but deeper and more tenacious. 2 For a long time it resisted their
encroachments. After the collapse of Mycenae it was approached
by wave upon wave of emigrants, but the ABolians settled
mostly north of the Hermos and the Dorians south of the
Maiandros. It was only when the tide was nearly spent that
the fertile middle region was transformed into Ionia.
Eratosthenes put the beginning of the ABolian migration at
1124 b.c., sixty years after the fall of Troy and eighty before
the Ionian migration, from which it is said to have differed in
being more desultory and protracted. 3 The original Aiolis,
according to Herodotus, comprised twelve towns: Killa in the
Troad; Pitane near the mouth of the Ka'ikos; Gryneia, Myrine,
Aigai, and Kyme, all on or near the coast to the south of Pitane;
Temnos in the hills overlooking the Hermos; in the lower
Hermos valley Larisa, Neonteichos, and Smyrna, all founded
from Kyme; still further south Notion on the coast near
Kolophon; and Aigiroessa, which has not been located. 4 Some
of these had been seized from the older inhabitants — Pelasgoi,
Carians, and Leleges in the coastal districts, and further in-
land two other peoples of the Caro-Lydian stock, the Mysoi
and Maiones. Herodotus seems to imply that these twelve
formed a sort of AEolic League, but there were others nearly as
old and recognised as AEolic — Tenedos, Lesbos, and Magnesia-
under-Sipylos.
In two cases we have some information about the manner in
which the settlements were made. This reaches us through
Strabo from Hellanikos,. a Lesbian antiquary of the fifth
century B.c. When the Dorians were overrunning the Pelopon-
nese, an expedition set out from Sparta under the leadership
of Orestes. He died in Arcadia, and was succeeded by his son,
Penthilos, who led the exiles as far as Mount Phrikion in
Lokris, where some of them remained. Penthilos continued his
journey overland to Thrace, where apparently he died. The
2 On the Hittite remains in this area see above p. 179. Miietos, Kolophon,
Erythrai, and Chios all claimed to have been founded originally from
Crete: Paus. 7. 2-4.
3 Str. 582. 4 Hdt. 1. 149.
XVII
THE HOMERIDAI
543
next stage was conducted by his son, Echelas, who, crossing
the Hellespont or Bosporos, pushed on as far as Daskylion; and
finally his youngest son, Gras, turned south and transported his
followers to Lesbos. 6 Meanwhile the party which had stayed
behind in Lokris sailed from Aulis under the command of
Kleuas and Malaos, also descended from Agamemnon, and
founded Kyme.°
The details of this tradition are open to question, but wo
of its implications may be taken as authentic and have a
bearing on the Homeric problem.
The first expedition, which seems to have taken a very long
time and to have had no clear idea of where it was going, has the
FIG. 84. Sub-Myeencan soldiers: vase from Tiryns
air of a really desperate adventure — an apt comment on the
poverty-stricken culture of the Peloponnese in die sub-
Mycencan period. 7 The second, which followed by sea, seems
to have been better organised. In both cases the majority of the
emigrants were probably drawn from Thessaly and Bceotia
(pp. 396-7). They are made to start from the Peloponnese simply
because their leaders were the exiled Pclopidai. This may be
6 Str. 582, cf. Hell. 1 14, Pi. N. 11. 3 4-5, Paus. 2. z8. 6, 3. 2. 1.
c Srr. 582, cf. 401. " Hall CGBA 239-86.
544 STUDIES IN ANCIENT GREEK SOCIETY XVII
accepted as a fact. We know that Mytilene, the chief town of
Lesbos, was ruled in early days by kings of the Perithilidai, 8
and at Kyme we hear of a king named Agamemnon, who
probably belonged to the other branch, represented by Kleuas
and Maiaos. 9 The aim of these fugitives, clinging loyally to the
last of the Pelopidai, was not to break with the past but to
transport it and preserve it in their new homes, which lay close
, to the scene of die Trojan War.
The Ionian colonisation was a much more vigorous affair.
It was directed by Neleidai exiled from Pylos to Athens. 10
There they had received grants of land and a place in the tribal
‘ system, which was perhaps reconstituted for the occasion. They
did well in Attica. One of their clans, the Medontidai, secured
the Athenian kingship; another, the Kodridai, led the migra-
tion to Ionia. Their part in the movement may be exaggerated
/in our tradition, which views the past through Athenian eyes,
but it must have been considerable. Some of the cities they
founded were organised on the basis of the four Attic tribes, 11
and all except two of them kept the Attic feast of the Apatouria. 1 2
The twelve cities of the Panionic League are divided by
Herodotus into four groups according to their dialects: (i)
Chios and Erythrai; (2) Ephesos, Kolophon, Lebedos, Teos,
Klazomenai, Phokaia; (3) Miletos, Myous, Priene; (4) Samos. 13
Four of these — Chios, Klazomenai, Phokaia, and Samos —
stand apart from the main movement. Chios was founded from
Euboia, Klazomenai from Kleonai and Phleious, Phokaia from
Phokis, Samos from Epidauros. 14 The last was brought into
the League forcibly by an expedition from Ephesos. 1 6 Phokaia
and Klazomenai were admitted after accepting Neleid kings,
the former from Erythrai and Teos, the latter from Kolophon. 1 6
All twelve were ruled at the outset by kings — KodridaL or
Glaukidai, or in some cases both. 17 The Glaukidai were a
Greek-speaking clan established for centuries at Xanthos, the
8 Arist. Pol. 1311b.
0 Poll. 9. 83. Near Smyrna there was a spring named after Agamemnon:
Philostr. Her. z. 18.
10 Hdt. x. 146, Paus. 7. 2. 1; see above pp. 390-2. 11 CIG. 3078, 3664.
12 Hdt. 1. 147. is Hdt. 1. 142. 14 Paus. 7. 3-4.
18 Paus. 7. 2. 8, 7. 4. 2. i° Paus. 7. 3. 10. 17 Hdt. 1. 147.
XVII
THE HOMERIDAI
545
capital of Lycia (p. 165). They may have been brought in
to conciliate the native population, which certainly survived at
Miletos and Tcos (p. 169) and probably everywhere. The
indigenous culture was too strong to be suppressed. The
official centre of the League was fixed at Panionion on Mount
Mykaie, but this was too far south to be convenient, and in
later times, when the Ionians had expanded in all directions,
they reunited at the festival of Apollo in Delos.
At Ephesos the Kodridai retained down to Roman times some
of their regal privileges, such as the right to wear purple and
the priesthood of Demeter Eleusinia. 18 How long the kingship
lasted in Ionia we do not know, but it probably declined
there more rapidly than in Aiolis.
The Ionians prospered. Though later in die field than die
^Eolians, they were soon strong enough to seize from diem all
their points of vantage. Chios was Ionicised at an early date
(p. 5 1 5). Smyrna was die next to go. It was well placed at the
mouth of die Hermos but hampered by Phokaia and Klazo-
menai at the entrance to die estuary, and before long it was
seized by an expeditionary force from Kolophon. The inhabi-
tants were permitted to withdraw to other parts of Aiolis, and
Smyrna became Ionian. 1 ® When the Hellespont was opened
up, the AEolians established themselves at Sestos and Abydos,
but later Abydos was annexed by Miletos, then Lampsakos
was founded from Phokaia, and the Milesians secured at
Kyzikos a foodiold sdll furdicr north on the Propontis.® 0
Again Aiolis had lost die lead, and she never recovered it.
With the exception of Mytilcnc, which secured a place in the
Greek concession at Naukratis in the Delta, 21 none of die
AEoIic setdements was able to stand up against die competi-
tion of Ionia.
„The early history of these colonies has long been familiar;
yet, fragmentary though it is, it contains some valuable clues
for the Homeric problem which have never been followed
up.
Epic poetry grows out of court minstrelsy, in which the king's
victories are commemorated. That is true everywhere, and
18 Str. 632. 10 Hdr. 1. 149-50, Paus. 7. 5. 1.
= ® J. L. Myres in CAH 3. 657-60. 21 Hdc. 2. 17S. 3.
Ll
546 STUDIES IN ANCIENT GREEK SOCIETY XVII
Greece is no exception. From a stray allusion in the Odyssey
we gather that Agamemnon’s minstrel at Mycenae had been
an official of high standing .® 2 When his descendants, the
Penthilidai, had finished their long trek, they recreated, so far
as their straitened circumstances permitted, their traditional
court life. The attempt was not a total failure, because, though
they had lost all their worldly capital, they had one cherished
heirloom which could not be taken from them. The very back-
wardness of Aiolis, by conserving ,the kingship, favoured the
cultivation of epic poetry.
The feats of Beowulf, recited at the Anglo-Saxon courts, had
been performed beyond the North Sea, and the hero can have
had no direct ties with the kings who listened to his adventures.
Neither Beowulf nor Widsith introduces any English character
with the single exception of Offa . 23 The Older Edda and the
Nibelutigenlied belong in their present form to Iceland and
Bavaria respectively; but their heroes, in so far as we can
identify their nationality, which is never stressed, are Goths,
Huns, and Burgundians . 24 It is a general characteristic of
Germanic epic, due to the extensive and protracted nature of
the migrations, which spread the Teutonic peoples over
nearly the whole of Europe, that the lays were preserved by
minstrels far removed in time and place from the persons and
events they commemorated.
These Penthilidai, on the other hand, had only crossed the
AEgean; and in their new home, within sight of Mount Ida,
which overlooked the battlefield, Agamemnon’s lineal de-
scendants listened to the Iliad. One can imagine them re-
marking to their guests, ‘It’s a small thing but our own’.
Then came the Kodridai. The story of Odysseus, whose
home was so close to their own ancestral seat, may well have
been their contribution to the Homeric treasury. His travels
in the west present some curious parallels with the voyage of
the Argonauts, leaving us with a suspicion that the saga may
have been transferred from east to west by the Neleidai when
they migrated from Iolkos . 2 5
In Ionia the kingship lasted just long enough to unite the
22 Od. 3. 267-71. 23 Chadwick HA 32.
24 lb. 33-4. 25 j, a. K. Thomson 80-99.
XVII
THE HOMERIDAI
547
two strands, and there the art was carried over without a break
into an atmosphere that no epic minstrel has ever breathed
before or since — the keen, critical, bracing air of tire mercantile
city-state.
2. Homer’s Birthplace
An enquiry into Homer’s birthplace can be undertaken
without prejudice to the question whether there was ever an
author of the Iliad and Odyssey in the ordinary sense of the
word. The Greeks believed there was, and they are entitled to
a hearing.
The Homeric Question is not a modern invention. Even in
the great days of Hellenistic scholarship it was being debated
whether or not the two poems had been written by the same
man. Such disputations flourished. The position reached in the
third century of our era is sketched by Lucian’s lively pen:
Two or three days later I met the poet Homer, and, since neither of us
was engaged, I took the opportunity of questioning him on various matters,
including his birthplace, which, as I explained, was still a subject of keen
controversy among us. He replied that he knew he was assigned by different
authorities to Chios, Smyrna, or Kolophon, but in reality he was a native of
Babylon, known to most people as Tigrancs; it was only after being sold to
the Greeks as a hostage (iSmtros) that he assumed the name Homer. I went
on to enquire whether the verses rejected by the editors were really his; he
replied that he had written them all. This prompted me to denounce all the
pedantic nonsense produced by the school of Aristarchos and Zcnodotos, and
after these points had been disposed of I asked him what his motive was in
beginning the Iliad with the wrath of Achilles. He replied that it was just
an idea that struck him, nothing more. I was also anxious to know whether
he had written the Odyssey first, as many authorities claimed, but he said no.
I had no need to ask him whether the story of his blindness was true, because
I could sec for myself that it was not . 20
That this gentle ridicule was not uncalled for can be seen from
the entry in which the Byzantine lexicographer Suidas sums
up die results of Homeric research. I quote the paragraph
referring to the poet’s birthplace:
Doubts whether a poet of such genius could have been mortal have led to
similar uncertainty in regard to his place of origin. He has been claimed by
various authorities as a native of Smyrna, Chios, Kolophon, Ios, Kyme,
Kenchreai in the Troad, Lydia, Athens, Ithaca, Cyprus, Salami's, Knossos,
Myccmc, Egypt, Thessaly, Italy, Lucania, Gryneia, Rome, and Rhodes. 5 "
50 Luc. VH. z. 20. Suid. 'Oanpoj.
548 STUDIES IN ANCIENT GREEK SOCIETY XVII
This is a formidable list of candidates. It was compiled,
however, in the eleventh century a.d. after nearly two mil-
lennia had been spent in pursuit of the truth. We may begin
by eliminating all those that cannot produce a pre-Christian
referee. This gives us a short list of seven:
Authority
Date
( century B.c.)
Birthplace
Homeric Hymn to Apollo
VII-VI
Chios
Semonides of Amorgos (?)
VH-VI
Chios
Simonides of Keos (?)
VI-V
Chios
Damastes of Sigeion
V
Chios
Pindar
V
f Chios
\ Smyrna
Stesimbrotos of Thasos
V
Smyrna
Hippias of Elis
V
Kyme
Bakchylides of Keos
V
los
Antimachos of Kolophon
v-rv
Kolophon
Ephoros of Kyme
IV
Kyme
Aristotle
IV
los
Philochoros of Athens
IV
Argos
Theokritos of Kos
m
Chios
Aristarchos of Samothraike
in-n
Athens
Nikandros of Kolophon
11
Kolophon
Dionysios Thraix
11
Athens 2 8
Let us interview these candidates, starting with the weakest.
Athens is soon dismissed. As the metropolis of the Ionians
she claimed the credit for their achievements — a claim which
in later times they accepted as a compliment to themselves.
Aristeides of Smyrna describes his city as an Athenian colony
and his forefathers as Athenians . 29 We possess an epigram com-
posed for a statue of the Athenian tyrant Peisistratos:
Thrice tyrant, thrice banished and restored, I am the statesman Peisis-
tratos, who collected Homer’s scattered lays; for, if Athenians founded
Smyrna, the golden poet was our fellow-countiyman. 3 °
los. There was a story that a girl of this island named Kre-
theis, got with child by a god, was sold into slavery at Smyrna,
where she was bought by a Lydian named Maion, who married
ss Horn. H. 3. 172, Sim. 85, Pi. fr. 264, Dam. io=FHG. 2. 66, Stesim.
18, Hippias &**FHG. 2. 62, B. fr. 48 Blass, Antim. 18 =FHG. 2. 58,
Eph. 164, Arist. fr. 66, Philoch. 54, Theoc. 7. 47, Nicand. fr. 14, VHom. 5-6.
20 Aristid. 23. 26, 29. 27, 40. 759, 42. 776.
so VHom. $- 6 =AP. 11. 442.
XVII THE HOMERIDAI 549
her, and she gave birth to Homer. 81 This was no doubt the
story to which Bakchylides and Aristotle referred, and it gives
Smyrna, not Ios, as the birthplace.
Argos. The Homeric poems were exceptionally popular here,
no doubt for political reasons. The Argives had a music
festival to which Homer and Apollo were invited as guests. 32
They said that Homer was a son of Maion and Hymetho. 33
Hymetho figures in this tradition as a variant of the girl from
Ios. She was the eponym of the Hymetheis, one of the Argive
tribes, composed of pre-Dorian elements (p. 166). It is not at
all impossible that some tradition of Mycenean minstrelsy, in-
dependent of the Homeric, survived here after the Dorian
conquest.
Kolophon. Being natives of the city, the sponsors of this
candidate are interested parties. The basis of their claim may
be that lonians from Kolophon repopulated Smyrna (p. 545).
We are left with Kyme, Smyrna, and Chios. Of these Kyme
is the weakest, and perhaps it only appears because it was the
metropolis of Smyrna. The strongest is Chios, supported by
the Homeric Hymn and Semonides of Amorgos — if it is he
and not Simonides of Keos to whom the citation refers. In
addition, Chios was the acknowledged home of the Homeridai. 34
In favour of Smyrna it might be conjectured that the Homeridai
transferred their centre from there to Chios after Smyrna fell to
Kolophon. Our best course is to declare both Chios and
Smyrna elected, with Kyme as proximo accessit. All three belong
to the borderland of Aiolis and Ionia, the coast of the Hermaic
Gulf. This was the cradle of Greek epic as we know it.
3 . From Court to Marht-place
Homeros is the eponym of the Homeridai. The name at
least is a real one. It is the Ionic form of Homaros, which
occurs as a. personal name in inscriptions from Crete and
Thessaly. 85 As a common noun it meant ‘hostage’, and there
was a story that the poet had been taken as a hostage from
31 PIu. VHom. 3. There was a month Homereon at Ios: IG. 12 (5) 15.
32 Ael. VH. 9. 15. 33 1 'Horn. 4. 1-2, 6. 27, Certamtn 25.
MStr. 645, Acus. 31, Hell. 55. 35 GDI. 1033, SIG. 1059. 1. 3.
55 ° STUDIES IN ANCIENT GREEK SOCIETY XVII
Smyrna to Chios. 36 The name explains the story. Another, more
plausible, account of it was that homeros was an old word
meaning ‘blind’. 37 Minstrels are often blind for the same
reason as smiths are lame. The choice of vocation was condi-
tioned by physical'infirmity. Blindness went with ‘second
sight’, that is, with prophecy and poetry. 38 Demodokos was
blind, so were Thamyris and Stesichoros. 39 If Homer was
simply ‘the blind bard’, his name does not argue much for his
reality.
Of the Homeridai we are told that ‘originally they had been
descendants of Homer who recited his poems by hereditary
tradition, but in later times they were rhapsodes unrelated to
the poet’. 40 In other words, they began as a clan and became a
guild. The qualification of birth was waived in favour of co-
option (p. 332). Their centre was Chios. Like all minstrels, they
were proverbial wanderers, and no doubt they had members in
many parts of Greece. One of them, Kynaithos of Chios,
migrated to Syracuse at the end of the sixth century. 41 They
were still flourishing in the fourth century, as we learn from
Plato, who mentions certain esoteric poems in their possession,
not available to the public. 42 By that time they may have
lost their monopoly of the poems, but it is remarkable that
nowhere in Attic literature nor in inscriptions do we find any
reference to a rhapsode who was an Athenian by birth. When
Plato wishes to portray a typical exponent of the art, he chooses
an Ionian from Ephesos.
If the poems matured at the courts of the Pelopidai and
Kodridai, the decline of the kingship must have affected them
decisively. Kyme had a king as late as 700 B.C., but this was an
extreme case. The office had probably been superseded, at
least in Ionia, long before that. It was this disparity of develop-
ment between different parts of Asiatic Greece that made it
possible for the epic tradition to be carried over into the next
stage without a break.
As the court declined, the recitals were transferred to the
30 Prod. Cbr. 99. 17 Allen. 3 ? lb. 19-20, Eph. 164.
38 Cf. Chadwick GL 3. 619. 33 Od. 8. 63-4, II. 2. 599-600, Iso. Hel. 64.
43 Pi. N. 2. 1 sch. 41 2. 1 sch.
42 PI. Phdr. 252b, cf. Io 53od, R. 599c.
XVII
THE HOMERIDAI
551
market-place. I do not mean a sleepy rural market, a rendezvous
for peasants, cattle-jobbers, and country squires. That is where
Hesiod recited, with results that can be measured by compar-
ing him with Homer. I mean the public square of some
populous sea-port, thronged with Greeks, Carians, Phoenicians,
merchant-seamen, textile manufacturers, moneylenders,
bankers, and above all the annual fair at Delos.
Delos is a tiny island, a mere outcrop of gneiss and granite
in the blue AEgean, but, set in the centre of the Cyclades, it
became the cultural metropolis of Ionia:
O Lord Apollo, many are thy shrines and wooded glades; all forelands and
mountain peaks are dear to thee, all rivers running seawards, but dearest of
all is Delos. There, trailing their long cloaks, the Ionians flock with their
wives and children to keep thy memory with boxing matches, dances, and
music — a sight so splendid that the onlooker, gazing in rapture at the throng
of men, women, ships, and merchandise, might think they were free from
old age and death. 43
And not from Ionia only — from all over Greece the pilgrims
flocked to the festival. Early in the eighth century we hear of
a chorus from Messenia competing with a hymn composed for
them by Eumelos of Corinth. 4 * Athenians were competing in
Solon’s time, and probably before . 45 In Greek, if you heard
someone singing very heartily, you said, ‘He is singing as
though he were bound for Delos '. 46 The island retained its pre-
eminence down to the Persian conquest, and after the defeat
of Persia its traditional prestige secured for it the treasury of
the new Ionian league formed by Athens.
That Homeric recitals were a prominent feature of the
programme is certain. The altar of the Delian Apollo is
mentioned in the Odyssey ,« and here, according to tradition,
the blind bard himself had once enthralled the crowds:
Well, may Apollo be gracious, and Artemis! Farewell, girls of Delos, and
remember me hereafter, when some distant traveller shall come and ask, ‘Of
all the minstrels that have visited you who has given the greatest delights'
— remember to answer with one voice, ‘A blind man, he dwells in rocky
Chios, and his songs shall never be surpassed.’ 48
43 Horn. H. 3. 143-55. 44 Paus. 4. 4. 1.
46 Ath. 234c, Philoch. 158. 46 Zen. 2. 37.
4 ? Oi. 6. 162-3, c ^' Cif rtamen 315-21, Hes. fr. 265.
48 Ho m. H. 3. 165-73.
552 STUDIES IN ANCIENT GREEK SOCIETY XVII
We do not know when the recitals were introduced there.
It may have been as far back as the ninth century. And, as we
learnt in the last chapter, the poems were still expanding in the
seventh. The change came before they were complete. It was
therefore formative. Indeed, it must have been revolutionary. .
They had grown up in the sheltered court life of an old-world
nobility, feeding on memories of the past. Now, thrown into
the hubbub of Ionian trade, politics, and science, they burst
into flower. The conditions were unique.
4. The Homeric Corpus
In the foregoing pages the Iliad , Odyssey, and Hymns have
been described collectively as the Homeric poems. In antiquity
there were about a dozen other works, now lost, current under
the name of Homer or his school. This was the Homeric
corpus. It falls into two portions. First, there were the poems
ascribed unanimously, or almost unanimously, to the master
himself— the Iliad, Odyssey, and Hymns. These I shall continue
to call the Homeric poems. The others, attributed variously
to him or his disciples, are known as the Cyclic poems.
Most of our information about the Cyclic poems conies from
Proklos the Neoplatonist (fifth century a.d.), who compiled a
guide to the Homeric corpus, of which a summary has sur-
vived. 40 He seems to have done his work thoroughly. Apart
from him, we have only quotations and allusions in other
writers and the fragments of Hellenistic, Graeco-Roman, and
Byzantine scholarship.
The Iliad deals with the tenth year of the Trojan War from
the quarrel between Agamemnon and Achilles to Hector’s
funeral. The subject of, the Odyssey is Odysseus’ return to
Ithaca, his reunion with his family, and his vengeance on the
suitors. The Iliad contains 15,693 verses, the Odyssey 12,110.
Both were divided by the Alexandrian editors into twenty-four
books. The Iliad was universally attributed to Homer; so was
the Odyssey, except that some Hellenistic scholars are said to
have dissented. 60
49 On the authorship of the Chrestomathia see Allen HOT 51-60.
so Prod. 102. 3 Allen.
XVII
THE HOMERIDAI
553
- Table XVIII
THE HOMERIC CORPUS
Title
Author
Date
Iliad
Odyssey
Homer
Homer
c. 950
Hymns:
Apollo
/Homer
(.Kynaithos of Chios
fl- 500
Others
Homer
Trojan Cycle:
Kypria
Aitliopis
Little Iliad
Sack of Troy
Homecomings
Tdcgonia
/Stasinos of Cyprus
LHegesinos of Salamis
Arktinos of Miletos
f Kinaithon of Sparta
j Lesches of Mytilene
1 Thestorides of Phokaia
[Diodoros of Erythrai
Arktinos of Miletos
Agias of Troizen
/Kinaithon of Sparta
lEugammon of Kyrene
b. 744
fl. 762
//. 710
b. 744
fl. 762
fl. 566
Thiban Cycle:
Oidipodeia
Thebais
Epigonoi
Kinaithon of Sparta
Homer
Antimachos of Teos
fl. 762
fl • 75 ?
Miscellaneous:
Capture of Oichalia
Battle of tie Titans
Plokais
Margites
Amazonia
Herdkltia
Kreophylos of Samos
/Arktinos of Miletos
IjEumelos of Corinth
Homer
Homer
Magnes of Lydia
Peisinos of Lindos
b. 744
fl. 75 o
fl. 700
fl. 750
Homer is mentioned only where there is no other candidate.
554 STUDIES IN ANCIENT GREEK SOCIETY XVII
There are thirty-four Hymns , but all except five of them are
very brief. The Apollo is quoted as Homer’s by Thucydides
(fifth century), the Hemes by Antigonos of Karystos (third
century). 61 Athehaios (second-third century a.d.) describes the
author of the Apollo as ’Homer or one of die Homeridai’. 62
Hippostratos of Syracuse (undated) says that the real author
of this hymn was one of the Homeridai, Kynaithos of Chios,
who ‘interpolated a great deal of his own verse into Homer’s’
and visited Syracuse between 504 and 500 B.c . 63 In recent years
Wade-Gety has .argued very forcibly that this poem consists
really of two hymns — one to the Delian Apollo, composed
before 600 b.c., and another to the Delphic Apollo, composed
during the next century, the combination being the- work of
Kynaithos . 64 1 accept this conclusion.
The Cyclic poems may be classified according to subject as
the Trojan Cycle, the Theban Cycle, and Miscellaneous.
There are six poems in the Trojan Cycle. First, the Kypria, in
eleven books. Its subject was the judgment of Paris, the rape
of Helen, the marshalling of the Achaeans, the sacrifice of
Iphigeneia, and the course of the war down to Agamemnon’s
quarrel with Achilles. 66 Herodotus (fifth century) argues from
internal evidence that Homer cannot have been the author,
implying that many people thought he was. 66 There was a
story, which can be traced to Pindar (fifth century) that Homer
gave the poem as a wedding present to his son-in-law, Stas-
inos of Cyprus. 67 Plato (fourth century) quotes from it without
naming an author. 68 Pausanias (second century a.d.) is also
non-committal. 69 Athenaios ascribes it to ‘Stasinos of Cyprus,
or Hegesias, or whoever he may have been’. 60 Proklos gives
him as Stasinos or Hegesinos of Salamis, i.e. the Cyprian
Salamis. 61
Second, the Aithiopis , in five books. Subject: the tenth year
of the war from after Hector’s funeral to the death of Achilles. 62
Proklos gives the author as Arktinos of Miletos, who is
61 Th. 3. 104, Antig. 7, cf. Paus. 4. 30. 4, 9. 30. 12, 10. 37. 5.
62 Ath. 22b. 68 Pi. N. 2. 1 sch. 64 Wade-Gery K. 66 Prod. 102-5.
66 Hdt. 2. 1 17. 67 Ael. VH. 9. 15, cf. Iamb. VP. 146, Said. 'Owipos29. -
68 PI. Euthyph. 1 2a. 60 Paus. 4. 2. 7.
so Ath. 682d, cf. 35c, 334b. 6i p r ocI. 97. 14.
02 Procl. 105-6.
XVII THE HOMERIDAI 555
described by Suidas (eleventh century a.d.) as a disciple of
Homer. 63 His date of birth is given as 744 B.c. 64
Third, the Little Iliad, in four books. Subject: the contest
for the armour of Achilles and the construction of the Wooden
Horse. 66 It is variously attributed to Kinaithon of Sparta
(Hellanikos, fifth century), Lesches of Mytilene (Proklos),
Thestorides of Phokaia, or Diodoros of Etythrai. 66 Kinaithon
was dated 762 B.c . 67 and Lesches was contemporary with
Arktinos. 68 There was a story that Homer had composed it
while staying at Phokaia with Thestorides. A son of Thes-
torides named Parthenios, also an epic poet, is described as a
descendant of Homer. 6 ® Pausanias treats the poem as
anonymous. 70
Fourth, the Sack of Troy, in two books, by Arktinos, author of
the Aithiopis , 71
Fifth, the Homecomings, in five books. Subject: the post-war
adventures of Diomedes, Nestor, Neoptolemos, Agamemnon,
and Menelaos. Author: Agias (Hegias) of Troizen (Proklos). 72
Pausanias mentions a poet of this name but treats the Home-
comings as anonymous. 73
Sixth, the Telegonia, in two books. Subject: the adventures
of Odysseus from after the funeral of the suitors to his death. 74
The author was Kinaithon of Sparta (Eusebios, third century
A.D.) or Eugammon of Kyrene (Clement, second-third century
a.d .). 76 Eugammon is dated 566 B.c.
The Trojan Cycle is discussed by Aristotle in terms which
show that he did not regard Homer as the author. 76
Next comes the Theban Cycle of three poems. The Oidi-
podeia told how CEdipus killed his father, married his mother,
and cursed his sons. The Thebais described the war between the
03 Suid. 'Apicrivos = FH G, 4. 314. 64 Said, l.c.: Allen HOT 62-3.
os Prod. 106-7. 66 E. Tr. 821 sch. 67 Allen HOT 63.
os Clem. Str. 1. 21. Lesches is quoted as having described (in a poem?)
a competition between Homer and Hesiod (Plu. M. 154a). The only point
in the story that need be mentioned here is that both poets are credited with
the faculty of improvisation.
0® Suid. TTapSivios. 70 Paus. 3. 26. 9. 71 Prod. 107-8.
72 lb. 108-9. 73 P aus> *• z ' I > IO< z 8* 7 . of. Ath. 281b.
74 Prod. 109. 76 Euseb. Chr. Ol. 4, Clem. Str. 6. 25. 1.
76 Arist. Poet. 23. 5-7.
556 STUDIES IN ANCIENT GREEK SOCIETY XVII
sons — the first Argive expedition against Thebes, ending in
their death at one another’s hands. The Epigonoi described die
destruction of the city in a second expedition organised by the
sons of the Argive leaders who had perished in the first. The
Thelais and the Epigonoi contained 7000 verses each. 77
The Oidipodeia is assigned in an inscription to Kinaithon. 78
Pausanias treats it as anonymous. 70 The Thebais was attributed
to Homer by Kallinos of Ephesos in the eighth century. 80 This .
is by far the earliest association of the master’s name with any
poem in the corpus. Herodotus refers to ’Homer’s Epigonoi , if
he is really the author of that work’. 81 In an Alexandrian
scholium it is attributed to Antimachos, probably Antimachos
of Teos ( fl . 753 b.c .). 82
There remain the miscellaneous works.
The Capture ofOichalia. Subject: the last exploit of Herakles.
Author: Kreophylos of Samos (Kallimachos, third century). 83
Plato mentions Kreophylos as ‘a friend of Homer’s'. 84 Else-
where he is described as his son-in-law, like Stasinos. 88
Kallimachos tells a story, which was probably known to Plato,
that Kreophylos received the poem as a gift from Homer after
entertaining him at Samos. 88 Clement says that it was stolen v
from Kreophylos by Panyasis of Halikarnassos. 87 Perhaps it
„ was not stolen but adapted.
The Battle of the Titans is attributed by Athenaios to Arktinos
of Miletos or Eumelos of Corinth. 88 Eumelos (fl. 750 b.c.)
belonged to the Bakchidai (p. 201). He was the reputed author
of another epic entitled the Korinthia , 88 It was he who com-
77 Ccrtamm 255-60, cf. CIG. It. Sic. 1292. 2. 12.
78 CIG. It. Sic. 1292. 2. xi. 78 Paus. 9. 5. 11.
80 Paus. 9. 9. 5. The restoration of KaWuvoj for KaXaTvos is virtually
certain, being supported by the accent as well as the common confusion of
A and A: Ko&alvoj is a vox nihili. Scott’s desperate plea that 'Opnpov may mean
'an Homer’ (UH 16) cannot stand: that would be SKKov 'Ounp°v or Safrspov
'Opilpov (see W. G. Hcadlam in G. Thomson AO 2. 93). Other writers treat
the Thebais as anonymous: Ath. 465c, Apld. 1. 8. 4.
81 Hdt. 4. 32. 88 Ar. Pa. 1270 sch., Plu. Rom. 12. 83 Call. Ep. 6.
84 PI. R. 600b. 88 Suid. KptobipuAoj. 88 Str. 638.
87 Clem. Str. 6. 25. 2. For the name cf. p. 167,
88 Ath. 22c, 277d, A.R. 1. 1195 sch., Hyg. F. 183, Clem. Str. 1. 21. 8.
80 Paus. 2. x. i, 2. 2. 2, 2. 3. 10.
XVII THE HOMERIDAI 557
posed the hymn for the Messenian competitors at Delos
55 i)'
The Phohais is said to have been taken from Homer by
Thestorides, who entertained him at Phokaia. 90 Nothing is
known of its contents.
The Margites was a burlesque in mixed hexameters and
trimeters about a simpleton who did not know which of his
parents had given birth to him and refused to make love to
his wife for fear she might tell his mother. 81 Plato and
Aristotle accept it as Homeric, but later writers treat it as
spurious. 92
Lastly, there are the Amazonia by Magnes of Lydia (fZ. 700
B.c.) and the Herakleia, which according to Clement was stolen
from Peisinos of Lindos by Peisandros of Kameiros (JL 750
B.C.). 93
Two more points, and our data will be complete. On the
one hand, Plato’s Ion, the rhapsode from Ephesos, announces
himself as a professional minstrel who specialises exclusively
in the works of Homer, 94 and all the quotations that follow
are from the Iliad and Odyssey. Similarly, Xenophon mentions
an Athenian who knew the whole of Homer by heart, meaning
by that, as the context shows, the Iliad and Odyssey . 96 On the
other hand, Proklos records that 'the ancients’ credited Homer
with all the Cyclic poems — that is, with the whole corpus. 96
We see that the ancient testimony is confused. What are
we to make of it? The treatment it has received hitherto
is, to say the least, capricious. In early times, according
to the separatists, all these poems had been attributed indis-
criminately to Homer, who was a mere eponym, devoid of
historical reality. The Unitarians, on the other hand, have been
at pains to show that, apart from the two masterpieces, none
of them was originally regarded as his. Both views have some
support in the data, which are contradictory, and therefore the
truth must be something different from either. The mistake
made by both schools is that they have tried to get rid of the
99 Ps. Hdt. VHom. 16. 91 Aeschin. Ct. 160. sch.
92 Pi. Ale. 2. i47c,'Arist. Poet. 4. 3. 10-2, Heph. Encheir. i7,DioPrus. 53. 4.
93 Nic. Dam. 62, Clem. Str. 6. 25. 2, Suid. TTelaovEpos.
94 Pi. lo 531a. 96 X. Sym. 3. 5, cf. Ath. 620b. 96 Prod. 102.
55^ STUDIES IN ANCIENT GREEK SOCIETY XVII
\
contradictions in the evidence instead of seizing on them as a
clue.
The Greek historian enjoys one great advantage. The
political separatism of the city-states favoured the survival of
parallel cults derived from a common original and alternative
versions of the same events/ offering copious material for re-
constructing the truth by comparison and analysis. The Greek
tradition is a tangled skein, which has to be unravelled by
identifying and following its separate threads. The strongest of
all was the Athenian, which after the fifth century tended to
gather the others into itself. But the Ionians had their own
culture, older than the Athenian, and it remained largely in-
dependent down to Hellenistic times. It has recently been
shown that ‘some of the Alexandrian scholars who came from
Ionia brought with them from their native cities a knowledge
of works which had never found their way to Athens at all ’. 97
Studied in this light, the contradictions in the Homeric tra-
dition can be resolved.
In the eighth century Kallinos, an Ionian, ascribes the
Thebais to Homer, and three centuries later Pindar tells the
story of Stasinos’ wedding present, implying that Homer was
the author of the Kypria. But then Hellanikos of Lesbos gives
the Little Iliad to Kinaithon, while Herodotus, a native of Asia
Minor who lived at Athens, feels it necessary to challenge the
view that the Kypria and Epigonoi were Homer’s. At Athens,
Thucydides quotes the Hymn to Apollo as Homeric, but in the
fourth century Xenophon excludes all save the Iliad and
Odyssey. Plato and Aristotle do the same except that they admit
the Margites. In the Alexandrian period, the names of several
rivals to Homer are known, and the general attitude is non-
committal.
This is not one tradition but two. Both of them developed,
and eventually they became entangled.
One was the tradition of the Homeridai themselves. In the
earliest times, when these ‘sons of Homer’ had been fellow
members of a real minstrel clan, they followed the pious
custom, common in such fraternities, of ascribing the whole of
their repertory to the master. Ipse dixit . Later, when the clan
97 Pearson 9 .
XVII THE HOMERIDAI 559
had become a guild, they were more individualistic. Being still
engaged in expanding or elaborating their inherited material,
they reconciled their personal claims with their esprit de corps
in anecdotes about wedding gifts and hospitable exchanges, in
which their own names and the master’s were symbolically
combined. In some cases the same theme was rehandled by
several of them in succession. In the conditions of oral recita-
tion this was natural and inevitable, but in later times, when
the claims of individual authorship had become paramount, it
led just as inevitably to misunderstanding. The successive
poets appeared as rivals guilty of interpolation or plagiarism.
As the poems became current on the mainland, the tendency
was at first to follow the earlier Ionian practice and treat all
alike as Homer’s, but in the fourth century, with the begin-
nings of literary criticism, Attic writers preferred to reserve
Homer’s name for the two chefs d'ceuvres together with the
Hymns and Margites, which had not been definitely individualised
even in Ionia. And finally the two traditions merged at
Alexandria. The names of Arktinos, Lesches, Kinaithon, and
the other Homeric poets, transmitted from Ionia, now became
generally familiar, but, owing to the influence of Attic litera-
ture, which ignored them, the attitude of educated people was
sceptical. Meanwhile the general public was quite content to
believe that Homer had written the whole corpus. If challenged,
they had only to appeal to his divine parentage.
5. The Cyclic Poems
Of the ten poets named in connection with the Trojan and
- Theban Cycles only five are described as natives of Aiolis or
Ionia. The remainder belong by birth or adoption to the Pelopon-
nese, Cyprus, or Libya. By testing their claims and the dates
assigned to them we may hope to discover something about
the expansion of the Homeridai.
Kinaithon of Sparta is dated 761-758 B.c. Even if we take
this as his date of birth, it is remarkably early — twenty years
before Lesches, his rival for the Little Iliad, and two hundred
before Eugammoh, his rival for the Telegonia. He cannot have
composed the Little Iliad in the form described by Proklos,
560 STUDIES IN ANCIENT GREEK SOCIETY XVII
because its subject shows that it must have been planned in
conjunction with the Aithiopis and the Sack of Troy , which
were the work of Arktinos; but there is no difficulty in ac-
cepting him as the author of an earlier'version. His Telegonia
may have provided a model for Eugammon in the same way.
Kyrene had been colonised from Thera, and Thera from
Sparta. 98 There remains the Oidipodeia, for which he is the only
candidate. In this connection his date agrees with that given
for Antimachos, author of the Epigonoi, and with the antiquity
of the Thelais, which was known to Kallinos. Kallinos lived
‘not long before* Archilochos, who has recently been dated
740-670 B.C."
It is thus quite probable that Kinaithon did belong to the
eighth century, and turning to Sparta we find him in con-
genial company. Not yet militarised, that city was then en-
joying a cultural renaissance, which attracted poets from all
parts of Greece — Thaletas from Crete (undated), Polymnastos
from Kolophon (undated), Terpandros from Lesbos (an old
man in 676 B.C.), Alkman from Sardeis ( fl . 672 or 657 b.c.), and
Tyrtaios from Athens (fl. 630 B.C.). Terpandros instituted
musical contests at the Kameia, 100 and Alkman must have been
familiar with the Odyssey , because he made a ballet of the ball-
game at which Odysseus surprised Nausikaa. 101 Further, the
Spartan lawgiver Lykourgos is said to have inaugurated re-
citals of the Iliad and Odyssey, which he had obtained from the
family of Kreophylos in Samos. 102 Lykourgos is an impalpable
figure, partly mythical, so we cannot give him a date, but this
story agrees with the rest in suggesting that the Homeridai
were patronised at Sparta in the eighth century. It may be
added that the legend of CEdipus had a special interest there.
He was one of the ancestors of the Spartan kings. 103
Agias of Troizen, author of the Homecomings , has no date.
Reference has already been made to the Homeric festival at
Argos and the story of Hyrnetho (p. 549). Down to the middle
of the seventh century die Argive kings rivalled the Spartan
08 Hdt. 4. 147-59. 08 Allen HOT 61, Blakeway DA.
100 Hell. 1 22. 101 Alcm. 16.
102 Plu. Lycurg. 4, Held. Pont. RP. 2. 3, Ael. VH. 13. 14.
103 Hdt. 6. 52. 2. Also of the Spartan Aigeidai: Hdt. 4. 149. 1.
XVII
THE HOMERIDAI
561
for the cultural leadership of the Peloponnese. The last of
them, Pheidon (Jl . 675 B.c.) seized control of the Olympic
Games. 10 * If there were already Homeric minstrels at the Spartan
court, they are likely to have found their way to Argos too.
Stasinos and Hegesinos, associated with the Kypria, are also
dateless. Both were Cypriotes, the latter from Salamis, which
was the royal seat of the Teukridai (p. 386). There was another
line of kings, the Kinyradai, at Paphos (p. 5 1 3). Both claimed
Achsean descent, and both lasted into Hellenistic times. These
must have been the patrons of Stasinos and Hegesinos. It is
even possible that Cyprus had an indigenous school of Achaean
minstrelsy, which merged with the Homeric.
Eugammon of Kyrene is dated securely at 566 B.c. Kyrene
was only founded in the last quarter of the preceding century.
Here too, trader the Battidai, the kingship persisted, and in
this case its connection with epic can be clinched. The
Telegonia was a sequel to the Odyssey. As the Greeks pushed
their way into the western Mediterranean, the saga of Odysseus,
being largely concerned with those regions, was expanded far
beyond its Homeric limits, and Odysseus became the father
of a large family. In particular, Eugammon gave Telemachos a
brother, Arkesilaos. 106 This was the name of at least four
Cyrenean kings. Evidently the Battidai claimed kin with
Odysseus. What the relationship was is uncertain, but it must
have been known to Eugammon, who in improving the
Odyssey furthered their interests.
Thus, the Homeridai established themselves at Sparta in
the eighth century, at Argos in the eighth or seventh, at
Kyrene at the end of the seventh, and in Cyprus some time in
the same period. Outside Ionia they found a congenial home
at the courts of kings.
In one of those brief but memorable paragraphs, of which
there are so many in his writings, Aristode contrasts the
Cydic poems with the Iliad:
Hence, as I hare remarked, it may well seem a sign of divine inspiration that
even in the Iliad, though the Trojan War has a beginning and an end, Homer
104 Hdt. 6. 127. 3. On his date see H. T. Wade-Gery in CAH 3. 761.
108 Eust. 1796. 50. The Battidai were descended from the Minyai: Hdt,
4. 150. 2, Pi. P. 4. 256-62.
Mm
5 & 2 , STUDIES IN ANCIENT GREEK SOCIETY XVII
does not set out to treat it as a whole. The subject was too long to be com-
prehended in one view, and, if he had tried to compress it, it would have
become too complicated. What he does is to concentrate on one portion,
which he diversifies with numerous episodes, like the Catalogue of Ships.
Other poets — for example, the authors of the Kypria and' the Little Iliad —
treat of a single character in a whole series of actions extended over a whole
period. And hence it is that, while neither the Iliad nor the Odyssey has
provided material for more than one or two tragedies, the Kypria has yielded
several, the Little Iliad eight or more. *00
fig. 85. King Arhsilas: Laconian cup
The Cyclic poets were inferior in constructive power. That was
the accepted opinion. Horace too contrasts them with Homer,
qui nil molitur inepte, and Proklos says they were studied mainly
for the interest of their subject-matter . 1 ' 17
Homer’s merit in Aristotle’s view was that he did not
100 Arist. Poet. 23. 5-7.
107 Hor. AP . 140, Procl. 97.
XVII the homeridai 563
attempt to compress his subject. The Cyclic poets did. The Iliad
and Odyssey both run to twenty-four books, though the action is
confined to a few weeks. The Kypria covered ten years in
eleven books; the Homecomings covered eight years in five books.
The scale was much smaller. And lastly, the Trojan Cycle
presupposes the Iliad and Odyssey in substantially their present
form. The Kypria ended where the Iliad begins; the Little Iliad •
began where the great Iliad ends; the Homecomings was a sup-
plement, and the Telegonia a sequel, to the Odyssey . The
creative power of the Homeridai had passed its peak.
We are now in a position to adumbrate three phases in the
history of the Homeric epos.
First, we have what may be called the primitive period of
short lays recited locally by court minstrels in Aiolis and
Ionia. This is the phase reflected in the lays of Phemios and
Demodokos. Beginning as one among many bardic clans, the
Homeridai built up a reputation as the outstanding exponents
of their craft. The Iliad and Odyssey were already taking shape,
but as loosely-strung sequences rather than organic wholes.
They had not yet crystallised.
Then the Homeridai secured a place in the Delian festival
of Apollo. Faced with new responsibilities, new opportunities,
they reorganise and expand. They abandon their exclusiveness
and become a professional corporation open to all minstrels
with the requisite qualifications. They absorb their rivals,
thereby enriching themselves. Their popularity is such that a
large share of the festival programme — probably several days —
is given up to them, and so they obtain an adequate setting for
the production of large-scale masterpieces. The technical skill
revealed in the structure of the Iliad and Odyssey presupposes a
high degree of external organisation. Accordingly we may
accept the conclusion implicit in the Hymn to Apollo that it
was at Delos, with all Ionia listening, that the blind bard's
disciples raised their art to a pitch of excellence never since
surpassed. The poems were still plastic and had not yet ceased
to expand, but it was here, recited in full year after year and
improved with each recital, that they were moulded, polished,
harmonised and unified.
In the third phase the art strikes roots beyond Ionia. As it
.564 STUDIES IN ANCIENT GREEK SOCIETY XVII
expands it declines. The minstrels are welcomed at Sparta,
Argos, Cyprus, Kyrene, but in these surroundings shorter lays
are in demand, and so they tend to become what their pre-
decessors had once been — verse chroniclers attached to the
courts of kings. The art ends by reversing the direction of its
growth. And gradually it ceases to be creative. In an age which
• has risen to new levels of material and intellectual life it is no
longer an adequate vehicle for historical narrative. The real
heir to Homer in the mature city-state is not the empty-
headed virtuoso described in Plato's Ion but the prose chronicler.
Like the rhapsodes, Herodotus used to recite in public, 108
and, though his medium was new, his technique of a central
theme, the Graeco-Persian War, diversified with geographical
and historical episodes, is essentially Homeric. The father of
history was a child of epic.
6. Diffusion of the Iliad and Odyssey
We are now at the crux of the Homeric Question. When
were the poems written down? The ancient tradition is quite
definite. After becoming current as scattered lays the Iliad
and Odyssey were collected and edited in their present form at
the end of the sixth century by the Athenian tyrants. Accord-
ingly the separatists have claimed that the poems are com-
pilations, not integral works of art. The Unitarians blankly
refuse to accept the evidence. Separatism has flourished
mainly in Germany, unitarianism in this country; and so
national antagonisms have added fuel to the odium philologicum.
My own position may be stated at once. The separatists are
right in accepting the evidence; the Unitarians are wrong in
permitting them to misinterpret it. I find myself in the com-
fortable if unfamiliar position of pleading for moderation
between extremes.
. Some scholars seem to assume that, sped on viewless wings,
the poems became universally familiar almost from the
moment they issued from the master’s mind. That is certainly
a mistake. Their diffusion was as uneven as the development
of the city-states. Further, it is obvious that they may have
108 Eus. Chr. Ol. 83 cf. Str. 18.
XVII
THE HOMERIDAI
565
been known to professional poets before becoming publicly
accessible, and they may have been recited in extracts before
becoming familiar as wholes. Let us begin therefore by asking
when, where, and how full public recitals of the Iliad and
Odyssey were instituted outside Ionia.
At Sparta and Argos they were already known in the eighth
and seventh centuries B.c. But there is a complication here.
These cultured Peloponnesian kingships did not last. At the
end of the seventh century, frightened by unrest among the
serfs, the Spartan landowners took charge of the monarchy and
turned the court into a barracks. There were no more poets in
the Vale of Eurotas . Meanwhile Argos had lost her commercial
supremacy to Corinth, whose situation on the Isthmus lay on
the direct route from the AEgean to -the Adriatic, which was
now being opened up. The Homeric recitals may have sur-
vived at Argos; at Sparta they did not.
By the close of the eighth century Corinth was already an
important shipbuilding centre, and it is then that the diffusion
of Corinthian pottery begins. 109 But her development was
peculiar. At Sparta the aristocracy seized power early enough
to prevent the growth of trade; at Corinth, thanks to her
favourable position, they were unable to do this, but they did
the next best thing. Under the Bakchidai they secured a
monopoly of it, which became a stranglehold. They were over-
thrown by Kypselos (p. 202), a merchant prince or tyrant
of the normal type (657 B.c.). Under him and his son Perian-
dros there was a commercial and cultural revival. It was Peri-
andros who patronised Arion, a poet from Lesbos. 110 And it was
in this period that Corinthian vase-painters began to depict
scenes from the Iliad with sufficient accuracy to argue direct
acquaintance with the poem. 111 We may presume that they got
their knowledge from public recitals, instituted by the
tyrants.
The first tyrant of Sikyon — also convenient to the Isthmus —
was Orthagoras, a contemporary of Kypselos. He too must
have patronised the minstrels; for we learn from Herodotus
- 199 H. T. Wade-Gery in CAH 3. 535, 539. 110 Hdt. 1. 23-4.
111 Johansen ITGK; J. D. Beazley in JHS 54. 85, Wade-Gery K77. I
have been unable to get hold of Johansen’s book.
566 STUDIES IN ANCIENT GREEK SOCIETY XVII
that half a century later his successor Kleisthenes ‘put an end
to the rhapsodic competitions at Sikyon because the Homeric,
poems were so full of praise for Argos and the Argives ’. 112
This was just after a war with Argos. It is not likely that the
ban lasted very long. As a former dependency of Agamemnon,
Sikyon was proud of her Ho leric tradition. Her antiquaries
claimed to have discovered a mistake in the Iliad . Gonoessa,
they said, which appears in our text as one of Agamemnon’s
demesnes near Sikyon, was a false reading for Donoessa.
They attributed the corruption to the Athenian editors . 113
Moving north into Bceotia, we are in a country with an inde-
pendent school of epic, and so the conditions are quite different.
Hesiod is undoubtedly a historical person, though not the
author of all his works. Herodotus regarded him as Homer’s
contemporary , 114 but his language is definitely post-Homeric,
and modem scholars assign him to the eighth century. He lived
at Askfa, a village near Thebes. It is not certain that he was
born there. He may have been brought there in childhood by
his father Dios, who was an immigrant from Kyme . 116 Here,
then, in an age when all crafts were commonly hereditary, we
have a professional minstrel whose father came from the very
district we have identified as the cradle of the Homeridai.
Was Dios one of them? The ancients held that he was Homer's
kinsman, and produced a pedigree . 110 This of course was a _
fiction, but experience has taught us not to despise it for that
reason. The content of the Hesiodic corpus is Boeotian, taken
from the choral poetry of prehistoric Thebes and Orchomenos,
but the form is purely Homeric. The Hesiodic dialect and
. the Hesiodic hexameter are identical with Homer’s, and this
can only mean that the Hesiodic school, as we know it, was
founded by a branch of the Homeridai.
112 Hdt. 5. 67. 113 Paus. 7. 26. 13.
114 Hdt. 2. 53. 2. The Certamen purports to be a competition between
the two, each being required to finish hexameters begun by the other.
Such competitions are mentioned in early Irish literature and survived
within living memory: see Hyde AD.
116 Hes. Op. 633-40, Certamen 51-2. Perhaps it is more likely that he
was bom in Bceotia, because the Aeolic form of his name appears to have
been AtatoSos 'auspicious journey’ (EM. 452. 37).
113 Certamen l.e., Procl. 100.
XVII
THE HOMERIDAI
567
How widely the Homeric poems were known in the Boeotia
of his day is another matter. He is said to have competed with
Homer, but in Chalkis and Delos, not in Boeotia. 117 There, and
6nly there, the Hesiodic school were able to hold their own.
They must have been conversant with the work of their rivals,
but they may well have been loth to include it in their public
performances. On the other hand, it is quite possible that they
transmitted it along with their own repertory to other profes-
sionals. 118 For these reasons we may regard Boeotia as a secondary
centre of diffusion.
Turning to the colonies beyond the Adriatic, we are faced
with the statement of Hippostratos that ‘the Homeric poems
were recited at Syracuse for the first time by Kynaithos in the
69th Olympiad’, i.e. between 504 and 500 B.c. 119 It was
Kynaithos who arranged the Hymn to Apollo (p. 554). Hippos-
tratos, be it noted, does not allege that the poems had been
previously unknown in this part of the world. They were
certainly accessible to Stesichoros (Jl. 692 B.C.), whose family
came from Lokris and claimed kin with Hesiod. 120 What
Hippostratos says is that this was the first recital, implying
that the poems were then given an official place in the Syracusan
calendar. And there is nothing improbable in that. It agrees
with such other evidence as we possess. One of the earliest
Homeric critics, Theagenes, was a native of Rhegion, and his
death may be placed in the last quarter of the sixth century. 121
Furthermore, when Kynaithos set foot in Syracuse, that city
was on the threshold of the most splendid epoch in her career.
The landed nobility were still in power, but the merchant-
class was rapidly maturing, and in the next generation the
tyrant Gelon refounded the city, built a new harbour, and
multiplied the population by enforced transfers from other
towns (485 B.c.). His court was to become the most brilliant
artistic centre in the west and a rival even to Athens at the
117 Certamen, Hes. fr. 265.
118 Eumelos of Corinth probably drew on the Hesiodic school; his
Titanomachia and Korinthia were both Hesiodic in subject.
118 Pi. N. 2. 1 sch.
120 Arist. fr. 524. Stesichoros' treatment of myths was largely Hesiodic: see
above p. 508 and cf. Philod. Piet. 24, Str. 42, Hes. Se. Arg., II. 15.333 sch.
121 Tat. Or. Cr. 31.
THE HOMERIDAI
XVII
568
height of her power. And so Syracuse reinforces die lesson we
have learnt from metropolitan Greece. Under the opulent
patronage of these merchant-princes the art of epic, which had
grown out of court life, came back into its own.
And now Athens. Peisistratos reigned from 540 to 527 B.c.
He was succeeded by his sons, Hipparchos and Hippias.
Hipparchos was assassinated in 514 B.c. and Hippias was ex-
pelled three years later. Thus the Athenian tyranny lasted a
bare thirty years, but its achievements were immense. The
Peisistratidai succeeded where others had failed. Polykrates,
the ambitious tyrant of Samos, had made a bid for the com-
mercial hegemony of the JBgean, and in pursuit of this objec-
tive he paid special attention to Delos. On his initiative the
adjacent island of Rheneia was consecrated to Apollo. 122 But
he was cut short in mid career by the Persian conquest of
Ionia. Peisistratos followed his example. He undertook a
purification of Delos itself, which he effected by clearing away
the graves on the land surrounding the temple. 1 23 His aim was to
enhance his prestige by securing the patronage of the great
Ionian festival. And his claim was an exceptionally strong one.
As a descendant of the Neleidai (p. 192) he was sprung from
the revered founders of Ionia, whose forefathers held an
honoured place in the Homeric poems. He was himself named
after Nestor’s youngest son, who accompanied Telemachos
from Pylos to Sparta. 124 One contribution of this remarkable
family to European civilisation is familiar. They founded the
art of tragedy. But modern scholars have been less appreciative
of their services to epic.
The Hipparchos is the title of one of the Platonic dialogues,
not -by Plato himself but by a disciple of his who lived in the
fourth century. Sokrates is talking to a friend:
It was Hipparchos, the son of Peisistratos from Philaidai, the eldest and
most cultured of his sons, who among many other brilliant achievements
introduced the Homeric poems into this country. He it was who made the
regulation, still in force, that the rhapsodes must recite the poems consecu-
tively according to the cue. He also sent for Anakreon from Teos, and
Simonides of Keos was constantly at his side, enjoying his munificence. He
did all this with the aim of educating his people. 126
122 Th. 3. 104. 2. 122 Th. 3. 104. x. 124 Hdt. 5. 65. 4.
126 Pi. Ilippareb. 228b, cf. Iso. 4. 159, Lycurg. Leo. 102.
XVII
THE HOMERIDAI
569
At Athens, as elsewhere, the tyranny ended in reaction, with
the result that it was condemned without discrimination by
the democrats who had overthrown it. It became the fashion to
transfer some of the tyrants’ reforms, including the regulation
just mentioned, to Solon, whom they regarded as the true
father of democracy. 120 But in the present case at least there is
no doubt where the credit really belongs. Here again we can
appeal to the irrefutable evidence of pottery. Scenes from the
Iliad occur on Attic vases as early as the second quarter of the
sixth century, but it is only in the last quarter — the time of
Hipparchos — that the painters show themselves to be
thoroughly familiar with the poem. 1 27
Is there the slightest reason for distrusting the conclusion on
which all these signs converge? Allen concedes that the state-
ment in the Hipparclos —
is a remarkable one to have been made not more than 150 years after the
supposed event. That the Homeric poems were previously unknown in
Greece is disproved by their diffusion and influence at Sikyon under
Kleisthenes; that they had already arrived at Athens appears from the appeal
made to them in the matter of Sigeion. ... It is singular that the historical
imagination of the later fourth century conceived an eposless Attica till the
time of the Peisistratidai. 128
There is so much that is singular in ancient Greece that her
modern historians sometimes find it difficult to keep their own
imaginations tinder control.
The matter of Sigeion was this. Some time in the sixth
century, under the Peisistratidai or earlier, Athens was in-
volved in a dispute with Mytilene for the possession of
Sigeion in the Troad, a keypoint for controlling the Hellespont,
and the Athenian spokesmen are said to have appealed to the
D.L. 1. 57. 127 Johansen ITG; see above n. 111.
128 Allen HOT zz8. He argued that 'the whole legend . . . was fab-
ricated by Megarian antiquaries’ (245) aggrieved by Homer’s treatment of
Salamis in the Catalogue (see below p. 572). He pointed out that Peisistratos
(like other tyrants) was included among the Seven Wise Men, and credited
with the works of Myson (D.L. 1. 13, 106-8, Aristox. 89); and that according
to Ar. Pa. 1071 sch. he was nicknamed Bakis (by a comedian?). From this he
deduced that there was 'as early as the fourth century what we may call a
Pisistratean mythology in existence, according to which he was a philosopher,
a writer under an assumed name, and an oracle-poet’ (247). It seems to me
that Allen is the myth-maker.
57 ° STUDIES IN ANCIENT GREEK SOCIETY XVII
Iliad to prove, in support of their claim, that Athenians had
fought in the Trojan War. 12 9 But there is nothing in the Hip-
parchos or any other ancient authority to justify the inference
that the poems ‘were previously unknown’. And the fact that
rhapsodic contests had been held at Sikyon early in the sixth
century is not a reason for doubting that they were instituted
at Athens several decades later.
Allen’s treatment of Kynaithos is even more unceremonious:
The date Ol. 69=504 B.c. is impossible, since Syracuse founded in 733
cannot have been without Homer for two hundred years, the internal allu-
sions and omissions in our hymn do not allow it to have been written at the
beginning of the fifth century, and Thucydides could not have quoted a
poem as Homeric which had been written less than fifty years before his
birth. Therefore the numeral is wrong. ... If Syracuse had heard Homer
for the first time in 504, how could the Athenian ambassador have quoted
the Catalogue to Gelon? Accordingly we rely on the anecdote and say that
Kynaithos lived and recited Homer at Syracuse soon after its settlement, i.e.
before 700 b.c. 18 ®
This from the scholar who hurls his Unitarian scorn at the
'rigmarole methodology’ of those separatist Germans. 181 The
separatists have certainly been guilty of dreadful blunders, but
Allen’s house is made of glass. Writing before Wade-Gery’s
analysis of the Hymn to Apollo , he may be excused on that
point, but, even if Kynaithos had composed the whole of it,
Thucydides would still have distinguished it from other
hymns to the same deity by giving it its conventional Homeric
title. As for the Athenian ambassador at Syracuse in 481 b.c., 182
he was quoting from a poem which had been publicly recited
at Athens for more than thirty years and at Syracuse for
twenty. To alter 500 to 700 by a mere flourish of the pen is a
bold move. The sole reason offered for it is that Syracuse
'cannot have been without Homer for two hundred years’.
Why not? Allen has no answer. He can only fall back on his
unsupported conviction that the poet’s text had been in
general circulation from the beginning. He assumes without
question that all poetry of the past conforms to the premisses of
contemporary literary criticism.
The Unitarians are afraid that, if they abandoned this posi-
tion, the gates would be opened to the enemy, who would
129 Hdt. 5. 94. 2. 130 Allen HOT65-6. wib.7. 182 Hdt.7. 161. 3.
XVII THE HOMERIDAI 571
break into the stronghold and cut their treasures to pieces.
Let me try to reassure them.
7. The Recension of Peisistratos
'Who', observes Cicero with his inevitable interrogative,
'was more learned, eloquent, and cultured in his age than
Peisistratos, by whom the works of Homer, previously con-
fused, are said to have been arranged in their present form?’ 133
Peisistratos is praised here on the same grounds as Hipparchos
in the dialogue. Cicero had studied at Athens, .and was
quoting an Athenian tradition.
The matter is mentioned again by Pausanias and TElian,
who add nothing new, and many centuries later in three
Byzantine scholia:
I. It is said that Peisistratos pieced together Homer’s poems, whose in-
ternal coherence had been disrupted by time, because they had been read at
random in scattered portions.
II. It is said that Homer’s poetry was perishing, because at that rime it was
transmitted by oral instruction, not in writing. In keeping with his noble
character the Athenian tyrant Peisistratos resolved to earn further admira-
tion by planning to record the poems in writing. He organised a public
competition, offering to everyone who knew the poems and could recite
them a prize of an obol a verse. In this way he collected all the readings and
handed them over to experts. [There follows the epigram quoted on p. 548.]
III. The Homeric poems, previously dispersed, were arranged in their
present form during the reign of Peisistratos by two scholars selected at the
time by Aristarchos and Zenodotos — not to be confused with the Ptolemaic
scholars of those names. Some authorities ascribe the Pisistratid recension to
four editors — Orpheus of Kroton, Zopyros of Herakleia, Onomakr itos of
Athens, and . . . [the last name is illegible]. 134
Here the tale has been embroidered with picturesque details for
the .edification of Byzantine schoolboys, but the central theme
is ' authentic. Onomakritos the Orphic, author of a poem
called Purifications, figures in Herodotus, who says he was
banished by Hipparchos for interpolating into an ancient
oracle some verses about Lemnos. 13 ® The motive is not clear,
but it was certainly political. A few years later (502-495 B.c.)
Lemnos became an Athenian dependency. 1 36
133 Cic. Or. 3. 137. 134 See Allen HOT 2.30-3.
138 Hdt. 7. 6. 3. 133 Hdt. 6. 140.
572 STUDIES IN ANCIENT GREEK SOCIETY XVII
Plutarch and TElian attribute a similar edition of the poems
to Lykourgos . 187 The stories about Lykourgos are suspect, and
this one may have been designed simply to provide the
Athenian recension with a Spartan parallel. But it is not im-
possible that something of the sort had been done in early
Sparta as a means of regulating the recitals.
In the Catalogue of Ships the contingent from Salamis is
described in two verses:
Ajax brought from Salamis twelve ships and stationed them where the
ranks of the Athenians were stationed. 188
The second verse is omitted in several MSS., including one of
the best, and in two papyri. It was condemned by the Alex-
andrian editors, as we learn from Strabo, who points out that
it is contradicted by several passages later in the poem . 139 In-
consistency is not of course a proof of composite authorship.
Even Homer nods. But this verse is suspect from another point
of view. It is evidently intended to imply that Salamis was an
Athenian dependency, or at least closely allied with Athens.
There is nothing in our other sources to suggest any connection
between the two communities at this early date. It seems
therefore to be an interpolation in the strict sense of the term.
The ancients recognised this, and knew where it came from.
One of Peisistratos’ achievements had been to annex the island
of Salamis, which had previously belonged to Megara, and he
inserted this verse in the Iliad to consolidate his title. All the
ancient authorities rejected it, including Aristotle , 140 and the
Megarians claimed to remember the verses, mentioning four
places in their territory, which Peisistratos had deleted . 141
The motive alleged is sufficient, but there may have been
another. As a native of Philaidai,' Peisistratos was a fellow
villager of the clansmen of that name, who were descended from
Philaios, a son of Ajax and an immigrant from Salamis (p. 121).
Their chief at this time was Miltiades, who migrated to the
Thracian side of the Hellespont, where he set himself up as a
tyrant ruling in the Athenian interest , 1 42 It was his nephew and
137 Plu. Lyctirg. 4, Ael. VH. 13. 14. 138 II. z. 558-9. 139 Str. 394.
14 0 Arist. Rl. I. 15, II. 3. 230 sch., Quintil. 5. 11. 40. Some made Solon
the interpolator: D. L. 1. 48, Plu. Sol. 10. ■
141 Str. 394. 142 Hdt. 6. 34-5.
XVII
THE HOMERIDAI
573
successor who conquered Lemnos. There is a thread running
through these events which has yet to be unravelled, but it
seems that the link which the tyrant fabricated between
Salamis and Athens may have been taken from a tradition
known to him from boyhood. Perhaps it was woven into the
poem by Ionian rhapsodes at his court, anxious to gratify their
patron.
Two more interpolations were attributed to Peisistratos — the
references to Theseus and to the immortality of Herakles. 143
The one was a tribute to the Athenian national hero, who as
such was post-Homeric (p. 264); the other was designed to
reconcile the death of Herakles with his divinity. If this was
the whole editorial fee, it was a very modest one.
Finally, we read in a scholium appended to the Lay of
Dolon (Iliad X):
They say that this lay was composed by Homer separately, not as part of
the Iliad, and that it was inserted here by Peisistratos. 144
For the separatists this is the hammer-blow that drives their
thesis home. The Lay of Dolon was merely the last of many
accretions.. The Iliad was a conglomeration. For the Unitarians
it is an acute embarrassment. They reject the tradition of the
Pisistratid recension in toto, not on any arguable grounds, but
simply because they refuse to believe it, and they point out,
quite correctly, that there is nothing in the language of Book X
to show that it is late.
Is this or any other book of the Iliad or Odyssey an interpola-
tion} No matter how long the Homeric controversy may con-
tinue to hum, this question will never find an answer. It is
meaningless. Like the Achseans and Trojans, the separatists
and Unitarians have been fighting for a phantom, and their
misdirected valour appears a 11 the more remarkable when we
find the truth of the matter stated succinctly in the ancient
sources which they have made their battleground.
It is easy to disparage the Byzantine scholiasts. They too had
their faults. They are sometimes extraordinarily obtuse. But
they spoke Greek. They had received in an unbroken line the
heritage of Greece. This was a precious asset. It enabled them
143 Plu. Thes. zo, Od. xi. 602 sch.
144 ll. 10 sch. ad. init.
574 STUDIES IN ANCIENT GREEK SOCIETY XVII
sometimes to pass on the truth without understanding it. I
learnt this when I was working on the text of AEschylus, and
my Homeric studies have confirmed it.
‘The internal coherence of the poems had been disrupted.*
We are not told that they had never possessed unity, but that
they had lost it. The Unitarians have not noticed this. ‘The
poems were perishing because they had been transmitted by
oral instruction, not in writing.* Wolf, repudiated on this
point even by his own followers, was right.
How long did it take to recite the Iliad ? All we can say is
that at the Athenian dramatic festivals four plays were performed
in a day. This would be less than half the number of verses
in the Iliad. The Odyssey is a little shorter. It seems unlikely that
the Iliad or Odyssey could have been recited in a day.
The poems matured in Ionia. When they were installed at
Delos, the programme was framed to accommodate them.
They were the most important items. The Ionians were pros-
perous, and could afford a lengthy festival.
Then they spread to Greece proper. There they were not at
home in the same way as they had been in Ionia. They had to
compete with local talent. Their length became a disadvantage.
The Cyclic poems, constructed on a smaller scale, were de-
signed to meet the new conditions. The Iliad and Odyssey were
recited, perhaps earlier and more widely than our records show,
but only in selections. They began to disintegrate.
Then came the revival. Everywhere along the trade routes the
enfeebled' aristocracy was challenged by energetic merchant
princes who had a direct interest in raising the material and
cultural standards of the people. The demand for Homer was
renewed, and in the reorganised festivals room was made for
the Iliad and Odyssey.
But this was not enough. The competing rhapsodes were still
offering the most popular pieces to the detriment of the whole.
Accordingly it was stipulated that they were to recite the poems
through in the proper order, each beginning where the last
left off. But t<3 make this regulation effective it was necessary to
know what the proper order was. The need arose for an official text.
It was a formidable undertaking, and fraught with an in-
herent difficulty. How were the Iliad and Odyssey to be defined?
XVII
THE HOMERIDAI-
575
Was the Lay of Dolon to be included in the Iliad) Was the
Odyssey to close with the reunion at the fireside? What about
the Deception of Zeus, which some said was irreverent, and
the Catalogue of Ships, which, referring properly to the out-
break of the war, did not really fit? On all these matters, not to
mention questions of phrasing, metre, and the digamma, prac-
tice varied not only from rhapsode to rhapsode but from one
performance by the same rhapsode to another. The poems were
still fluid. Peisistratos was faced with the complex and delicate
task of recording and arranging a copious, richly diversified,
organic mass of oral tradition. The measure of his success is
the Iliad and Odyssey.
8. The End of Epic
The Greek alphabet was constructed in Ionia and diffused
by trade. In each community the spread of literacy was neces-
sarily slow. The initiative was taken by the merchants, who
wanted an instrument for commercial contracts and codifying
the laws. The landowners resisted for this reason. And
naturally it made less headway in professions wedded to an oral
technique.
The power of memory characteristic of preliterate peoples is
astonishing only to those who have not experienced it. Being
the only medium for preserving knowledge, it has been made
perfect by practice. Minstrels in particular have raised it to the
highest pitch. It is part of their craft. This explains why Greek
epic was so long in being committed to writing. The Homeridai
had no use for the pen. They carried their repertory in their
heads. The result was that, when they had diffused it beyond
their power to control it, they came very near to losing it. It
was saved by the merchant princes. History was kind to them —
how kind can be seen from what has happened elsewhere in
analogous conditions.
The peculiar beauty of epic diction, as compared with
written poetry, is its fluency and freshness. That is the virtue
of improvisation. It takes on new colours as it passes from one
festive occasion to another, sparkles in response to each
momentary stimulus. But. its lustre cannot be caught. Its
words are winged and cannot be pinned down.
576 STUDIES IN ANCIENT GREEK SOCIETY XVII
Let us take another lesson from the Kirghiz. Radlov de-
scribes his efforts to record their poetry:
Noting down the songs from dictation was very difficult. Not being
accustomed to speak slowly enough to be followed by the pen, the singer
loses the thread of the narrative and by omissions falls into contradictions.
... In spite of all my efforts I have not succeeded in reproducing their
minstrel poetry completely. The repeated singing of the same song, the slow
dictation, and my frequent interruptions dispelled the excitement in-
dispensable for good singing. The minstrel could only dictate in a tired and
negligent way what he had delivered before with fire. . . . Tfie verses written
down have therefore lost their freshness. 146
The heroic verse of most primitive peoples has suffered' ir-
reparable loss. Not only has much of it perished but what
survives has been mutilated. This has been shown by Soviet
research. The construction of some of these oral .epics, when
recited in the proper manner and environment, is faultless;
it is only in print that they manifest the discrepancies and con-
fusions often regarded as characteristic of popular poetry . 340
In the Soviet Union, however, the difficulty has been solved.
Not only have the minstrels been taught to write in circum-
stances that enhance their pride in' their national traditions,
but they have been equipped with the phonograph and
radio. These songs will survive. They have been saved by
machinery . 147
But these conditions are unprecedented. Elsewhere and at
other times the transition from speech to writing has been left
to chance. The best Germanic epics contain many fine things,
and, if they are inferior to Homer, this is largely due to losses
in transmission. The spread of literacy during the so-called
Dark Ages is thus described by Chadwick:
Three phases are to be distinguished in the early history of Roman writing
among the Teutonic peoples. In the first phase only Latin is written. In the
second the native language is employed for writing religious and other works
derived from Roman sources or based on Roman models. In the third phase
purely native works are written. But this third phase did not arise on the
continent before the twelfth century, and then only in a much modified
form, while even the second phase was largely local and hardly recognised in
the highest circles. 148
146 Radlov PV 5. xv. 148 Zazubrin, quoted by Chadwick GL 3. 180.
14 7 G. Thomson MP 56-8. 148 Chadwick GL 1. 483.
XVII
THE HOMERIDAI
577
In western Europe, after a violent upheaval, the heroic king-
ship was consolidated into feudalism, and when, after several
centuries, the power of the feudal lords was broken by the
bourgeoisie, the art of minstrelsy was dead. In Greece the ex-
pansion of trade was so rapid that the merchant-class was able
to appropriate die epic tradition in its prime. In western
Europe writing was introduced through an alien medium
which was an exclusive instrument of the ruling class. In
Greece the alien languages had been absorbed. In western
Europe popular poetry, being pagan, was suppressed. ‘When
priests dine together*, wrote Alcuin to the Bishop of Lindis-
fame, 'let the words of God be read. It is fitting on such occa-
sions to listen to a reader, not a harper, to the discourses of the
Fathers, not the poems of the heathen. What has Ingeld to do
with Christ?' 140 The Greek rhapsode was an honoured guest at
court and a repository of sacred lore. At every turn he had the
advantage over the scop, thanks to the extremely rapid and un-
even development of Greek society.
Who then is Homer? He is not a compilation. The separatists
made the mistake of leaving out his poetry. Nor is he a solitary
miracle. The Unitarians do not want to explain Homer, but
to envelop him in the magic of individuality and the miracle
of genius. But, though his songs have never been surpassed,
they are not a miracle. Homer is not one but many hereditary
poets, gifted and practised, who, together with the enthusiastic
crowds that spurred them to excd themselves and the far-
sighted statesman that saved their masterpieces for posterity,
may be described in Shelley's words as both creations and
creators of their age.
9. Structure of the Iliad and Odyssey
Before concluding let me outline briefly how the two poems
seem to me to have been built up. It must be brief, because a
detailed exposition lies beyond the scope of this work. And of
course.it is purely conjectural.
The kernel of the Iliad was a sordid quarrel between a chief
from the Thessalian backwoods and his Mycenean overlord.
140 Id. HA 41.
Nn
578 STUDIES IN ANCIENT GREEK SOCIETY XVII
Achilles refused to fight; the Achaeans were pressed back; his
thane, Patroklos, went to their assistance, and fell to Hector;
then, stirred to action, Achilles killed Hector, outraged the
corpse, and flung dogs, horses, men on his friend’s burning
body in an insensate orgy of revenge.
Tliis theme was developed, as Aristotle says, by means of
episodes. The first was the rejected offer of amends (Book IX).
This led to others. The Achaeans were defeated twice, before and
after the refusal, and the battle scenes were expanded further
by introducing the exploits of Diomedes and Agamemnon. But
above all, by refusing Agamemnon’s offer, Achilles took on
himself the responsibility for the quarrel, and so a new con-
clusion was called for to mark his change of heart. At this point
the minstrels took leave of the saga, treating it with the
imaginative freedom of conscious art. Deaf to all appeals from
his wife and family, the grief-stricken old king seeks out the
man who has killed his son and begs him for the body. The
two enemies pour out their hearts in a flood of tears, die one
for his son and the other for his father. The conflict is resolved.
The pathos of this climax was heightened further by episodes
from the life of the doomed city, introducing us to Helen and
Hector’s infant son. And over all hangs the mirage of the gods,
whose quarrels always end in laughter because they are im-
mortal.
This was a work of centuries. The Mycenean monarchy rose
and fell while the poem was being composed. The sophisticated
artists who added the finishing touches were far removed from
the semi-barbarous brigands of whom they sang. The result
was a dynamic tension between them and their material, and
they had absorbed their material so deeply that the tension
appears as something internal in their characters. ‘If’, says
Sarpedon to his vassal, ‘we were destined to live for ever like
the gods and never grow old or die, I should not send you into
battle nor would I go myself; but since in any case we are sur-
rounded by a thousand deaths and dangers, let us go — to get
glory or to give it .’ 1 B0 That is not the voice of a robber chief. The
Achilles who drew his sword on the king, sulked in his tent,
sobbed like a child, spurned the offer of cities, rolled in the
1B » il , 12. 322-8.
XVII
THE HOMERIDAI
579
dust for grief, and dragged his enemy’s body at die tailpiece
of his chariot — that is the authentic Achaean, the turbulent
catde-raider and pillager of Knossos. But Achilles is doomed;
so is Agamemnon, and Ajax. Their empire is nothing but a
memory conjured up out of the past in the magical hexa-
meters of poets who love to note the movement of sheep
stampeding in the fold, the sweep of a scythe in the grass, or
the grace of a woman’s fingers at the loom. And so, as
they see him, Achilles is tormented by foreknowledge of his
fate: ‘Shall I go home to Phthia and live out my life in un-
eventful ease, or die young in batde and live for ever on the
lips of minstrelsy?’ 151 The tragic dilemma of the Iliad crystallises
five centuries of revolutionary change.
The Odyssey contains a much larger admixture of non-
heroic fiction. It belongs more completely to the maturity of
the art, and so makes a smoother, more effortless unity.
In its present form it falls into six sections marking the pro-
gress of the action. Telemachos leaves home to seek news of his
'father (Books I-IV). Meanwhile Odysseus has landed in
Phteacia (V-VUI), where he tells the story of his wanderings
(IX-XII). Returning independendy, the father and son meet
in the swineherd’s hut (XUI-XVI). Disguised as a beggar and
insulted in his own house, Odysseus interviews Penelope and
prepares his plot (XVII-XX); and finally after slaying the
suitors he discloses himself to his wife and father and is re-
stored to peaceful possession of his heritage (XXI-XXIV).
The nucleus was a cunning man’s voyage overseas among
miracles and monsters and his revenge on the enemies who took
advantage of his absence. This is crude folklore, far older than
Odysseus. But in the poem, though the adventures extend over
ten years, all save the last are concentrated in a single section.
As in the Iliad, a large subject is comprehended in one view by
fixing the focus on a single portion, but here the method seems
to have been applied more consciously, and it has been carried
so far that the hero’s adventures are in effect subordinated to a
new interest — the reunion of his family. This is done by en-
veloping die main theme in four ancillary episodes.
First, the voyage of Telcmachos sets the stage in Ithaca and
151 Cf. ll. 9. 412-6.
580 STUDIES IN ANCIENT GREEK SOCIETY XVII
introduces the situation dramatically. His journey to Pylos
and Sparta is a heroic subject, but even in our text there are
signs of an older version in which after leaving Sparta he had
gone on to Crete. 162 It looks as though a separate lay has been
incorporated and adapted.
Secondly, Odysseus’ sojourn in the exotic island of Phaeacia
is a genuine reminiscence of a matriarchal Minoan city-state
(pp. 418-20) designed as a contrast to the rough island home
which he refuses to exchange for it. This too may once have
existed as a separate lay. The last of many adventures, it has
been transformed with superb skill into a lens through which
we view the rest. The one-eyed Cyclops and Circe’s witchcraft,
the twittering ghosts of hell and the fatal music of the Sirens
reach us as an enchanting tale at the fireside, twice removed
from reality.
Thirdly, the scenes in the swineherd’s hovel, which are pure
fiction, seem to have been suggested by the lay of Telemachos,
who was forced to return secretly. Their dramatic effectiveness
speaks for itself, but a word must be said about the swineherd.
Thersites in the Iliad is a man of straw, a butt of class prejudice,
but Eumaios is drawn with Shakespearean sympathy:
O good old man, how well in thee appears
The constant service of the antique world,
When service sweat for duty, not for meedl
The swineherd’s hut has no place in the heroic age. It is feudal,
and I suspect that even the feudal world of this gentle, gracious
peasant was already antique for the poets who imagined him.
Lastly, the slaughter of the suitors, though part of the
original theme, has probably been modified. The archery
contest, founded on the svayamvara (p. 404), is older than its
present context, and it is noteworthy that none of the suitors
has any independent place in genealogy or myth. 1 6 3 The brutality
of the scene conforms to the law of patriarchal Greece, which
permitted a man taken in adultery to be killed with impunity.
For Odysseus reunion with his family is synonymous with
recovery of his property.
162 Od. 1. 93, 284, 3. '313 sch.
103 With the possible exception of Peisandros son of Polyktor: see above
' p.426.
XVII THE HOMERIDAI 581
These episodes are fused with the original theme so skilfully
that they enclose the fabulous kernel in the much more human
story of what was happening in Ithaca. We meet his wife and
family before we meet him, and when we meet him he is all
but home again. The enchantments of Circe and Calypso only
sharpen his yearning for the sight of a whiff of smoke curling
up from the chimneys he knows so well. His nostalgia holds
us in suspense, and it is enhanced by a Leitmotiv — a device
which is not used in the Iliad. Right from the opening of Book
I our hopes and fears for the family have been played on by
repeated parallels with the fate of Agamemnon, who had a
fair wind home only to be murdered by his wife and her
paramour. Will Penelope keep faith? Can a hundred suitors
fail where Aigisthos succeeded? Will Telemachos prove, like
Orestes, a son worthy of his father? In Phaeacia this motive is
treated as a scherzo in Aphrodite’s dalliance with Ares during
the absence of Hephaistos. 164 Again, as in the Iliad, human
tragedy is divine comedy. And it culminates in the last book,
where we hear the ghost of Agamemnon pronounce Odysseus
happy. This book may be later than the rest, but it is justified.
Not only is the reunion with Laertes necessary — it is for this
alone the old man has lived so long — but the scene in which he
cries out for proof and Odysseus counts die trees he had
helped him to plant in the garden when he was a boy, is as
moving as any in Homer. At the end of the story the grand-
father, father, and son face the world together. There were
further adventures in store for Odysseus, but the Odyssey ends
there, and righdy.
All this, however, is litde more than guesswork. We are not
in a position to identify the raw materials. But at least we can
appreciate the technique. All theories of authorship, single or
composite, are beside the point. The concept of authorship is
inapplicable. These poems took shape out of a kaleidoscopic
background of impromptu variations adjusted to the inspira-
tion of the moment, crystallising gradually as the power of
improvisation failed. And dicy were brought to rest so gendy
diat in dicir final configuration die simple realism and natural
1 64 This point has been overlooked by those who reject the lay of Dcmodokos
as an ‘interpolation’.
582 STUDIES IN ANCIENT GREEK SOCIETY XVII
eloquence of primitive, popular poetry was combined with the
subtle, self-critical individualism of mature art. That is their
unique quality. From the nature of the case they could not
have been produced either by a single artist or by a succession
of artists working separately for their own ends. They were the
work of a school in which generations of disciplined and de-
voted masters and pupils had given their lives to perfecting
their inheritance. And all this was rendered possible by a
unique combination of historical circumstances, which laid a
bridge between improvisation and composition, between
speech and writing, so that something of the unpremeditated
audacity of the primitive minstrel, inspired by the shining
eyes and breathless silence of the crowd, was carried over into
the impassive but durable medium of the written word.
BIBLIOGRAPHY
Where an English translation of a foreign work is specified, the references
are to its pages and not to those of the original. A list of periodicals,
with the abbreviations used for them, will be found at the end of the
Bibliography.
Adam, L. Primitive Art. London, 1940.
Albright, W. F. ‘The Phoenician Inscriptions of the Tenth Century B.C.
from Byblos.' JAO 67. 153.
Allen, T. W. Homer: Origins and Transmission. Oxford, 1924.
The Homeric Catalogue of Ships. London, 1921.
Andersen, J. C. Maori Life in Ao-tea. Wellington, N.Z., 1907.
Ansted, D. T. The Ionian Islands. London, 1863.
ANTONIADIS, S. Place de la liturgie dans la tradition des lettres grecques. Leiden,
1939 *
Ashburner, W. ‘The Farmer’s Law.’ JHS 30. 85, 32. 68, 35. 76.
Bachofen, J. J. Das Mutterrecht. Stuttgart, 1861.
Baden-Powell, B. H. The Indian Village Community. London, 1896.
Baldwin Brown, G. The Art of the Cave-Dweller. London, 1928.
Bancroft, H. H. Native Races of the Pacific States of North America. London,
1875-6.
BASEDOW, H. The Australian Aboriginal. Adelaide, 1929.
BATESON, G. Navan. Cambridge, 1936.
Bateson, W. Letters Jrom the Steppe. London, 1928.
Baumeister, A. Denhmaler des klassischen Altertums. Munich/Lcipzig, 1889.
Beauchet, L. Histoire dtt droit privl de la rlpubliquc atllnienne. Paris, 1897.
Bellows, H. A. The Poetic Edda. London, 1923.
BeNVENISTE, A. ‘Lc sens du mot koAocotSs.’ RP 58. 118.
Bergk, T. Ueber das alteste Versmaass der Criechen. Freiburg, 1854.
Bishop, C. W. ‘Beginnings of Civilisation in Eastern Asia.' An 14. 301.
Blakeway, A. A. ‘The Date of Archilochus.’ Greek Poetry and Life (Oxford,
1936), 34.
Blegen, C. W. 'The Coming of the Greeks.' AJA 32. 146.
AND Kouroniotis, K. ‘Excavations at Pylos, 1939.’ AJA 43. 557.
Bleichsteiner, R. ‘Die kaukasischc Sprachgruppc.' As 32. 61.
Bleuler, E. Lehrbuch der Psychiatric. 3 cd. Berlin, 1920.
Bloch, M. ‘The Rise of Dependent Cultivation and Scignorial Institutions.’
Cambridge Economic History, Vol. I, 1941.
Blumel, R. ‘Homerisch vcrpxOso.’ Gl 25. 78.
Boas, F. Primitive Art. Oslo, 1927.
Boisacq, E. Dicticnnsire /tymclogique de la langue grecque. 5 cd. Paris, 1933.
Bonwick, J. Daily Life er.d Origin of the Tasmanians. London, J S70.
Bossert, H. T. 'Die Bcschworung ciner Krankhcit in der Sprache von
Kreta.’ OL 34. 303.
584 STUDIES IN ANCIENT GREEK SOCIETY
Bourke, J. G. The Snake Dance of the Moquis of Arizona. London, 1884.
Bowra, C. M. 'Homeric Words in Arcadian Inscriptions.' CQ 20. 168.
Tradition and Design in the Iliad. Oxford, 1929.
Brbal, M. Essai de slmantique. Paris 1899.
Breasted, J. H. History of Egypt. 2 ed. New York, 1910.
Briffault, R. The Mothers. London, 1927.
Bruck, E. F. Totenteil and Seelgerat im griech'ischen Recht. Munich, 1926.
BRUGSCH, H. Religion und Mytho\ogie der alten Aegypter. Leipzig, 1885.
Bucher, K. Arbeit und Rhythmus. 5 ed. Leipzig, 1919.
Buck, C. D. Comparative Grammar of Greek and Latin. Chicago, 1933.
Greek Dialects. 2 ed. Boston, 1928.
Budge, E. A. W. The Gods of the Egyptians. London, 1904.
History of Egypt. London, 1902.
Osiris and the Egyptian Resurrection. New York, 1911.
Buradkar, M. P. ‘Clan Organisation of the Gonds.' MI 27. 127.
Burkitt, M C. Prehistory. 2 ed. Cambridge, 1925.
South Africa's Past in Stone and Paint. Cambridge, 1928.
Burn, A. R. ‘Dates in Early Greek History.’ JHS 55. 130.
Minoans , Philistines and Greeks. London, 1930.
BURROWS, R. M. The Discoveries in Crete. London, 1907.
BURTON, R. F. The Lake Regions of Central Africa. London, i860.
Bury, J. B. History of Greece. 2 ed. London, 1913.
Calhoun, G. M. Growth of Criminal Law in Ancient Greece. Berkeley, 1927.
Cambridge Ancient History. Cambridge, 1925-39.
Cambridge History of India. Cambridge, 1922-32.
Cameron, A. L. P. 'Some Tribes of New South Wales.' JAI 14. 351.
Casteret, N. Ten Years under the Earth. London, 1940.
Caudwell, C. Illusion and Reality. 2 ed. London, 1946.
Cauer, P. Gnmdfrage der Homerkritik. 3 ed. Berlin, 1923.
CAVAIGNAC, E. Leproblime hittite. Paris, 1936.
Chadwick, H. M. The Growth of Literature. Cambridge, 1925-39.
The Heroic Age. Cambridge, 1912.
Origin of the English People. Cambridge, 1906.
CHANTRAINE, P. Grammaire homerique. Paris, 1942.
Morphologic historique dugrec. Paris, 1945.
Childe, V. G. ‘Archeology and Anthropology.' SJA 2. 343.
The Aryans. London, 1926.
'The Date and Origin of Minyan Ware.' JHS 35. 196.
The Dawn of European Civilisation. 3 ed. London, 1939.
Man Makes Himself. London, 1936.
Scotland Before the Scots. London, 1946.
CLARK, G. From Savagery to Civilisation. London, 1946.
Clay, R. The Tenure of Land in Babylonia and Assyria. London, 1938.
CodRINGTON, R. H. The Melanesians. Oxford, 1891.
Colutz, H. AND Bechtel, F. Sammlung der griechischen Dialektinschriftcn.
Gottingen, 1884-1915.
BIBLIOGRAPHY
585
Cook, A. B. ‘Who was the Wife of Zeus? CR 20. 365, 416.
Zeus. Cambridge 1914-40.
‘Zeus, Jupiter and the Oak’. CR 17. 174.
Cornford, F. M. ‘The ’Airapxcrf and the Eleusinian Mysteries.’ Quiggin 153.
From Religion to Philosophy. London, 1913.
CoRTSEN, S. P. ‘Die Lemnische Inschrift.’ Gl 18. 101.
CUNY, A. 'Le nom des ioniens.' RHA 7. 21.
CuQ, E. Etudes sur le droit babylonien, Its loss assyriennes et les lois hittites. Paris,
1929.
CuREAU, A. L. Savage Man in Central Africa. London, 1915.
Czaplicka, M. A. Aboriginal Siberia. Oxford, 1914.
Dange, S. A. Land Fragments and our Farmer. Bombay, 1947.
DaREMBERG, C. and Saguo, E. Dictionnaire des antiquitis grecques et romaints.
Paris, 1877-1919.
DARESTE, R., Haussouiller, B., and Reinach, T. Rtctteil des inscriptions
juridiques grecques. Paris, 1891-8.
Das, J. K. ‘Notes on the Economic and Agricultural Life of a Little-known
Tribe on the Eastern Frontier of India.’ As 32. 440.
DAWKINS, R. M. The Sanctuary of Artemis Ortlia at Sparta. London, 1929.
Deborin, A. M. and others. Bottpocbt ucmopuu doruiaccosozo oSiqecmea.
Moscow/Leningrad, 1936.
De Cara, P. C. A. Gli hcthei-pelasgi: ricerche di storia e di arclcologia orientate ,
greca ed italica. Rome, 1894-1902.
DeMETRAKOS, D. Mfya Atfnciv tt)S &At|VHCt|s yAdxnrns. Athens, 1956-.
De MORGAN, J. Prehistoric Man. London, 1924.
DePradennE, A. V. La prehistoirc, Paris, 1938. Prehistory. London, 1940.
Deubner, L. Attische Feste. Berlin, 1932.
Diamond, A. S. Primitive Law. London, 1935.
DlEHL, E. Anthologia Lyrica Graeca. Leipzig, 1925.
Dillon, M. The Cycles of the Kings. Oxford, 1946.
DlNNEEN, P. Filidke na Mdighe. Dublin, 1 906.
Dorsey, J. O. ‘Siouan Sociolog) - .’ ARB 15.
DR1BERG, J. H. The Lango: a Nilotic Tribe of Uganda. London, 1923.
Durham, M. E. Tribal Origins , Lew and Customs of the Balkans. London,
* 934 -
Durkheim, E., and Mauss, M. ‘Dc quclques formes primitives de classi-
fication.’ AS 6.
Dutt, R. P. India To-day. London, 1940.
Earthy, E. D. Valer.g: Il'cmm. Oxford, 1933.
Eggan, F. Social Anthropology of North American Tribes. Chicago, 1937.
Ehrenberg, V. People of Aristophsr.es. Oxford, 1943.
EHRENFELS, O. R. Mother-right in India. Oxford, 1941.
ElSLER, R. Orphan the Fisher. London, 1921.
‘Die Sccvolkcmamen in den altorientalischen Quellcn.’ C 5. 73.
Elderkin, G. W. ‘The Marriage of Zeus and Hera.' AJA 41. 424.
Elderkin, K. M. ‘Jointed Dolls in Antiquit}-.’ AJA 34. 455.
586 STUDIES IN ANCIENT GREEK SOCIETY
Elkin, A. P. 'Rock Paintings of North-West Australia.’ O I. 257.
Elliot Smith, G. Evolution of Man. Oxford, 1924.
Engberg, R. M. The Hyhos Reconsidered. Chicago, 1937.
Engels, F. Dialektik der Natur. (Marx-Engels Gesamtausgabe.) Dialectics
of Nature. London, 1940.
Ludwig Feuerlach. (Marx-Engels Gesam,tausgabe). London, n.d.
Der Ursprtmg der Familie , des Frivateigentums und des Staats. (Marx-
Engels Gesamtausgabe.) Origin of the Family , Private Property and the State.
London, 1940.
Engnell, I. Studies in the Divine Kingship in the Ancient Near East. Uppsala,
1945.
Entwhistle, W J. European Balladry. Oxford, 1939.
Erdmann, W. Die Ehe im alten Griechenland. Munich, 1934.
Ernout, A., and Meillet, A. Dictionnaire Itymologique de la langue latine.
Paris, 1932.
Esmein, A. ‘La propri£t£ fonjifere dans les pofemes homeriques.’ NRH 14. 821.
Evans, A. J. ‘Knossos Excavations, 1903.' ABS 9.
‘The Mycenean Tree and Pillar Cult.' JHS 21. 99.
The Palace of Minos. London, 1921-35.
‘The Ring of Nestor.’ JHS 45. 1.
Shift Graves and Beehive Tombs of Mycenn. London, 1929.
Evans, I. H. N. The Negritos of Malaya. Cambridge, 1937.
FARNELL, L. R. Cults of the Greek States. Oxford, 1896-1909.
Greek Hero Cults. Oxford, 1921.
Fick, A. Die Entstehung der Odyssee. Gottingen, 1910.
Die homerische Odyssee in der ursprunglichen Sprachform wiederhergestellt.
Gottingen, 1883.
Die llias in der urspringlicben Sprachform wiederhergestellt. Gottingen, 1886.
Firth, R. We, the Tikopia. London, 1936.
Fison, L. ‘The Nanga.’ JAI 14. 14.
Forrer, E. ‘Fur die Griechen in den Boghazkoi-Inschriften.’ KF 1. 252.
Forsdyke, E. J. ‘The Pottery called Minyan Ware.' JHS 34. 126.
Fortune, R. The Sorcerers of Dobu. London, 1932.
Fotheringham, J. K. 'Cleostratus.' JHS 39. 164.
Fowler, W.W. ‘Mundus Patet.' JRS 39. 164.
Roman Festivals. London, 1899.
Fox, C. E. The Threshold of the Pacific. London, 1924.
Fox, H. M. ‘Lunar Periodicity in Reproduction.' PRS. B. 95. 526.
Selene, or Sex and the Moon. London, 1928.
FRANKEL, H. Die homerische Gleichnisse. Gottingen, 1921.
Frazer, J. G. Apollodorus. London, 1921.
Folklore and the Old Testament. London, 1919.
The Golden Bough. London, 1923-7. The Magic Art and the Evolution of
Kings. Taboo and the Perils of the Soul. The Dying God. Adonis, Attis,
Osiris. Spirits of the Corn and of the Wild. The Scapegoat. Balder the Beautiful,
j the Fire Festivals of Europe, and the Doctrine if the External Soul.
BIBLIOGRAPHY 587
Frazer, J. G. Lectures on the Early History of the Kingship. London, 1905.
Pausanias's Description of Greece. London, 1898.
Totemica. London, 1937*
Totemism and Exogamy. London, 1910.
FroDIN, O. and Persson, A. W. Asine. Stockholm, 1938.
FORTWANGLER, A., AND Rbichhold, C. Griechische Vasenmalerei. Munich,
1904-32.
Fustel de Coulanges. La cite antique. 7 ed. Paris, 1878.
Gardiner, E. N. Olympia , its History and Remains. Oxford, 1925.
Garstang, J. The Hittite Empire. London, 1929.
Gaster, T. H. ‘The Graces in Semitic Folklore.’ JRA 1938. 37.
• 'Ras Shamra, 1929-39.' An 13. 304.
Gatschet, A. S. A Migration Legend of the Creek Indians. Philadelphia, 1884.
Gerhard, E. Auserlesene griechische Vasenbilder. Berlin, 1840.
Glanville, S. R. The Legacy'of Egypt. Oxford, 1942.
Glotz, G. La cite grecque. Paris, 1928.
La civilisation Iglenne. Paris. 1923. JEgean Civilisation. London, 1925.
I« solidariti de lafamille dans le droit criminel en Grice. Paris, 1904.
Le travail dans la Grice ancienne. Paris, 1920. Ancient Greece at Work.
London, 1926.
Goldenweiser, A. Anthropology. London, 1937.
Granet, M. La civilisation clinoise. Paris, 1929. Chinese Civilisation. London,
I 93°-
Gray, H. L. English Field Systems. Cambridge, Mass., 1915.
Gray, L. H. Foundations of Language. New York, 1939.
Grey, G. Journals of Two Expeditions of Discovery in North-Western and Western
Australia. London, 1841.
GrGnbech, V. Culture of the Teutons. Oxford, 1931.
Grote, G. History of Greece. 2 ed. London, 1869.
Grundy, G. B. Thucydides and the History if His Age. London, 1911.
Gruppe, O. Griechische Mythologie und Religionsgeschichte. Munich, 1906.
Gugushviu, A. 'Nicholas Marr.' Ge 1. 101.
Guiraud, P. La propriltc foneilrt en Grice. Paris, 1893.
Gummere, F. B. Old English Ballads. Boston, 1894.
Gunn, D. L., Jenkin, P. M., and Gunn, A. L. ‘Menstrual Periodicity.'
JOG 44. 839.
GuRDON, P. R. T. The Khasis. London, 1 914.
Haddon, A. C. Evolution in Art. London, 1895.
Reports of the Cambridge Anthropological Expedition to the Torres Straits.
Cambridge, 1908.
Hadow, W. H. A Comparison of Poetry and Music. Cambridge, 1926.
Haeckel, J. 'Totemismus und Zweiklassensystem bei den Sioux-Indianern.
As 32. 210, 450.
Hale, H. The Iroquois Book of Rites. Philadelphia, 1883.
Haley, J. B. ‘The Coming of the Greeks.' AJA 32. 141.
588 STUDIES IN ANCIENT GREEK SOCIETY
Hall, H. R. Ancient History of the Near East. 7 ed. London, 1927.
‘The Caucasian Relations of the Peoples of the Sea/ K 22. 335.
The Civilisation of Greece in the Bronze Age. London, 1928.
‘The Keftiu and the Peoples of the Sea.’ ABS 8. 157.
Halliday, W. R. ‘The Hybristika.' ABS 16. 212.
Hambly, W. D. Origins of Education among Primitive Peoples. London, 1926.
Hammond, J. L. and B. The Village Labourer. 4 ed. London, 1936.
Hansen, H. D. Early Civilisation in Thessaly. Baltimore, 1933.
Harrison, J. E. Prolegomena to the Study of Greek Religion. 3 ed. Cambridge,
1922.
‘Primitive Hero Worship.’ CR 6. 474, 7. 74.
‘Sophocles’ Ichneutse and the Dromena of Kyllerife and the Satyrs/
Quiggin 136.
Themis. Cambridge, 1912.
HARRISON, T. Savage Civilisation. London, 1937.
Hasebroek, J. Staat und Handel in alten Griechenland. Tubingen, 1928. Trade
and Politics in Ancient Greece. London, 1933.
HASTINGS, J. Encyclopedia of Ethics and Religion. Edinburgh, 1908-18.
Hawes, H. B., Williams, R. E., Seager, R. B., and Hall, E. H.
Gournia, Vasiliki, and other Prehistoric Sites on the Isthmus of Hierapetra
(Crete). Philadelphia, 1908.
HAWKES, C. F. C. Prehistoric Foundations of Europe. London, 1 93 9.
Head, B. V. Historia Numorum. 2 ed. Oxford, 1911.
HeicHELHEIM, F. Wirtschaftsgeschichte des Altertums. Leiden, 1939.
Heitland, W. E. Agricola. Cambridge, 1921.
HERTER, H. ‘Theseus der Athener.’ RM 88. 244, 289.
Herzfeld, E. E. Archeological History of Iran. London, 1935.
HEURTLEY, W. A. Prehistoric Macedonia. Cambridge, 1939.
'The Site of the Palace of Odysseus.' An 9. 410.
Hobhouse, L. T., Wheeler, G. C., and Ginsberg, T. Material Culture
and Social Institutions of the Simpler Peoples. London, 1930.
Hocart, A. M. Kingship. Oxford, 1927.
Hoebel, E. A. 'Comanche and Shoshone Relationship Systems.’ AA41. 440.
Hogarth, D. G. Excavations at Ephesus: the Archaic Artemision. London, 1908.
Hollis, A. C. The Masai, their Language and Folklore. Oxford, 1905.
The Nandi , their Language and Folklore. Oxford, 1 909.
Homolle, T. ‘Comptes des hieropes du temple d'Apoilon d£lien/ BCH 6. 1.
Hooke, S. H. Myth and Ritual. Oxford, 1 93 3.
Origins of Early Semitic Ritual. London, 1938.
Hopkins, C. ‘The Early History of Greece/ YCS 2. 1 15.
Hort, A. Theophrastus , Enquiry into Plants. London, 1916.
Hose, C., and McDougall, W. Pagan Tribes of Borneo. London, 1912.
Howitt, A. W. ‘Native Tribes of South-East Australia.' JAI 37. 268.
Native Tribes of South-East Australia. London, 1904.
Hubert, H. Greatness and Decline of the Celts. London, 1934.
Hunt, A. S. Tragicorum Grcecorum fragmentapapyracea. Oxford, 1912.
BIBLIOGRAPHY
589
Hutton, J. H. The Serna Nagas. London, 1921.
Hyde, D. Abhrdin diadha ClAige Clonnacht. London, 1906.
Imhoof-Blumer, F. ‘Coin Types of Kilikian Cities.' JHS 18. 161.
and Gardner, P. 'Numismatic Commentary on Pausanias.' JHS 7, 99.
IvENS, W. G. Island Builders of the Pacific. London, 1930.
The Melanesians of the South-East Solomon Islands. London, 1927.
Jackson, K. Early Celtic Nature Poetry. Cambridge, 1935.
Jarde, A. La formation dtt peuple grec. Paris, 1923. Formation of the Greek
People. London, 1926.
Jeanmaire, H. Couroi et Courltes. Liiie, 1939.
Jeanroy, A. iff 0 rigines de la pofsie lyrique. 2 ed. Paris, 1904.
Jebb. R. C. Homer. 7 ed. Glasgow, 1905.
Johansen, K. F. Iliaden i tidlig graesk kunst. Copenhagen, 1934.
Jolowicz, H. F. Historical Introduction to the Study of Roman Law. Cambridge,
1939.
Joyce, P. W. Social History of Ancient Ireland. London, 1903.
Junod, H. A. The Bantu Heritage. Johannesburg, 1938.
Life of a South African Tribe. 2 ed. London, 1927.
Kagarov, E. a&p. 3 menhc u npoSneMa puMCKozo poda.» Deborin 619.
nepeoKumKU nepeoSumHozo K0MMyHU3Ma e oSufecmeeHHOM empoe
dpeBHUx zpenoe u zepMaHtfee. Moscow/Leningrad, 1937.
KaRSTBN, R. The Civilisation of the South American Indians. London, 1926.
King, L. W. ‘Sennacherib ana the Ionians,’ JHS 30. 327.
Kirchhoff, A. Homers Odyssee. Berlin, 1859.
KlRCHNER, J. Prosopographia Attica. Berlin, 1901.
KOnig, F. W. ‘Mutterecht und Thronfolge im alten Elam' Festschrift der
National-hibliothek in Wien. Vienna, 1926.
Relief und Inschrift des Konigs Dareios von Bagistan. Leiden, 1938.
Die Stele von Xanthos: Metrik und Inhalt. Vienna, 1936.
Koschaker, P. ‘Fratriarchat, Hausgemeinschaft und Mutterrecht in Keil-
inschriften.' ZA 41. 1.
Kovalevsky, M. M. Tableau des origines de V /volution de la famille et de la
propriitL Stockholm, 1890.
KRAPPE, A. H. ‘’Att 6 WiCcv KOkvos.’ HS 37. 353.
Krause, W. 'Die Ausdriicke fiCir das Schicksal bei Homer.' Gl 25. 143.
Kretschmer, P. ' Alaksandus von Vilusa.’ Gl 18. 205.
'Zum altesten Sprachgeschichte Kleinasiens.’ Gl 21. 76.
Einleitung %ttr Geschichte der griechischen Sprache. Gottingen, 1896.
‘Etrusker und Pelasger.’ Gl 11. 276.
‘Zur Frage der griechischen Namen in den hethitischen Texten.' Gl
18. 161.
‘Zur Geschichte der griechischen Dialekte.' Gl I. 9.
‘Die griechische Benennung des Bruders.' Gl 2. 201.
‘Die Hypachaer.’ Gl 21. 213.
‘Mythische Namen.’ Gl 8. izi.
‘Der Name der Lykier und anderekleinasiatische Volkemamen.' KF 1. 1.
590 STUDIES IN ANCIENT GREEK SOCIETY
Kretschmer, P. ‘Nochmal die Hypachaer und Alaksandus.* Gl 24. 203.
‘Die Steliung der lykischen Sprache.’ Gl 27. 256, 28. 101.
Krichevski, E. uCucmeMbi podcmea, xax ucmoHHUK pexoHcmpyxyuu
pa 38 umuH coifitfiAbHou opeaHuaaifUu aecmpa/iuficxux nneMeH.v
Deborin 257.
Krige, E. J. Social System oj the Zulus. London, 1936.
Kroeber, A. L. ‘The Classificatory System of Relationship.' JAI 39. 77.
Kuhner, R., Blass, F., and Gerth, B. Ausjtihrliche Grammatik der grie-
chischen Sprache. Hanover, 1890-1908.
Kuiper, F. ‘Beitrage zur griechischen Etymologie.’ Gl 21. 267.
Lachmann, K. Betrachtungen iiber Homers Ilias. Berlin, 1847.
Lambrino, M. F. Les vases archaiques d’Histria. Bucharest, 1938.
Landtman, G. Origin of the Inequality of the Social Classes. London, 1938.
Lang, R. H. ‘Archaic Survivals in Cyprus.’ JAI 12. 186.
Langdon, S. The Babylonian Epic of Creation. Oxford, 1923.
Babylonian Menologies and Semitic Calendars. London, 1935*
Laroche, E. ‘Recherches sur les noms des dieux hittites.’ RHA 7.
Latischew, B. 'Zur Epigraphik von Bootien und Lamia.' MDA 7. 349.
Layard, J. Stone Men of Malekula. London, 1942.
Leaf, W. Troy. Oxford, 1912.
LEAKEY, L. S. B. Stone Age Africa. Oxford, 1936.
Legrand, E. ‘Inscriptions de Tr^zfcne.’ BCH 16. 84.
LejeunB, M. Traite de phonltique grecque. Paris, 1947.
LENIN, V. I. Selected Works. London, n.d.
Lethaby, W. R. ‘The Earlier Temple of Artemis at Ephesus.' JHS 37. I.
Lin Yueh-Hwa. ‘Kinship System of the Lolo.' HJA 9. 81.
Lobeck, C. A. Aglaophamus. Koriigsberg, 1825.
Lodge, O. Peasant Life in Jugoslavia. London, 1941.
Lolling, H. G. 'Aufgrabung am Artemision auf Nord-Euboia.' MDA 8.
200.
'Inschriften von Hellespont.’ MDA 9. 60.
‘Mittheilungen aus Kleinasien.’ MDA 7. 1 5 1 .
Lorimer, H. L. ‘Homer’s Use of the Past.’ JHS 49. 145.
‘Pulvis et Umbra.’ JHS 53. 161.
Lowie, R. H. Primitive Society. New York, 1929.
Luquet, G. H. ‘Sur les mutilations digitales.' JP 35. 548.
Macauster, R. A. S. Textbook of European Archeology, Cambridge, 1921.
MacCurdy, J. T. The Psychology of Emotion. London, 1925.
Macleod, J. The New Soviet Theatre. London, 1943.
Macpherson, S. Form in Music. London, 1915.
Maine, H. S. Village Communities in the East and West. 2 ed. London, 1890.
Malinowski, B. Coral Gardens and Their Magic. London, 1935.
The Family among the Australian Aborigines. London, 1913.
‘Kinship.’ M 30. 19.
'The Problem of Meaning in Primitive Languages.’ Ogden 296.
— — Sex and Repression in Savage Society. London, 1927.
BIBLIOGRAPHY 59I
Man, E. H./On the Aboriginal Inhabitants of the Andaman Islands.’ JAI
12. 327.
Mannhardt, W. Gtrmanischt Mythen. Berlin, 1858.
Markwart, J. 'Woher stammt der Name Kaukasus?’ C 6. 25.
Marr, N. H<pemunecKua Kaenaa u mpemuu smumecKuit sneMenm e
C03udaHUU cpeduseMHOMopcKOii Ky/ibmypbt. Leipzig, 1920. Der
japhetitische Kaukasus und das iritte Element im Bildungsproyess der
mittellandischen Kultur. Berlin, 1923.
■ and Prijuzhin, A. G. npoS/teMbi ucmopuu doKanumanucmmeaatx.
■ftopMatfUii. Moscow/Leningrad, 1934.
Marx, K. Das Kapital. (Marx-Engels Gesamtausgabe.) Capital. Vol. I.
London, 1938. Vols. II, HI. Chicago, 1933.
AND Engels, F. Briefwecbsel. (Marx-Engels Gesamtausgabe.) Selected
Correspondence . London, 1934.
Mason, O. T. Woman's Share in Primitive Culture. London, 1895.
Mathew, J. Two Representative Tribes of Queensland. London, 1910.
Matsumoto, H. "Hie Stone Age People of Japan.' AA 23. 50.
McKenzie, D. The Infancy of Medicine. London, 1927.
Meade, M. Co-operation ana Competition among Primitive Peoples. New York,
1937 .
MEEK, C. K. A Sudanese Kingdom: an Ethnographical Study of the Jukun-speaking
Peoples of Nigeria. London, 1931.
MehxeT, A. Aperpu d’une histoire de la langue grecque. 4 ed. Paris, 1935.
Introiiiirtion a I'ltude comparative its langues indo-europfennes. 8 ed. Paris,
1937 .
and Vendryes, J. Grammaire comparle des langues classifies. 2 ed. Paris,
1927.
Meissner, B. Babylonien und Assyrien. Heidelberg, 192.5.
Meister, K. Die homerische Kunstsprache. (Preisschriften der Jablonowskischen
Gesellschaft zu Leipzig 48.) 1921.
Melas, S. *0 Nwiapxos Miao^rAris. Athens, 1932.
Meyer, E. Forschungen zur alien Geschichte. Halle, 1892-9.
• Geschichte des Altertums. 2 ed. Stuttgart, 1937.
- Michell, H. Economics of Ancient Greece . Cambridge, 1940.
Michbll, J. and R. The Russians in Central Asia. London, 1865.
Mills, J. P. The Ao Nagas. London, 1926.
; The Rengma Nagas. London, 1927.
Mommsen, A. Feste der Stadt Atlen. Leipzig, 1898.
Mommsen, T. Romische Geschichte. Berlin, 1881-5. History of Rome. London,
1868.
Moorhouse, A. C. ’The Name of the Euxine Pontus.' CQ 34. 123.
Moret, A., and Davy, G. Des clans aux empires. Paris, 1923. From Tribe to
Empire. London, 1926.
Morgan, L. H. Ancient Society. 2 ed. Chicago, 1910.
The Hague of the Iroquois. Rochester, 1851.
Systems of Consanguinity and Affinity of the Human Family. New York, 1871.
592 - STUDIES IN ANCIENT GREEK SOCIETY
Muller, K. O. Orclomenos und die Minyer. z cd. Breslau, 1844.
Prolegomena %it einer wissenschaftlichen Mythologie. Gottingen, 1825.
Munro, J. A. R. ‘Pelasgians and Ionians.’ JHS 54. 140.
Murko, M. ‘Neues fiber sfidslavisches Volksepik.' NJK 43. 273.
Murr, J. Die Pflanzenwelt in der grieehisehen Mythologie . Innsbruck, 1890.
MYLONAS, G. E. ‘*EAEucnvicn« 5 c.’ AJA 40. 415.
Myres, J. L. History of Rome. London, 1914.
Who Were the Greeks? Berkeley, 1930.
Niebuhr, B. G. Romische Geschichte. Berlin, 1811-2. History of Rome. London,
1885.
Nilsson, M. P. ‘Die Entstehung und Religiose Bedeutung des grieehisehen
Kalenders.' Lunds Universitets Arsskrift, N.F., Avd. I, 14. 21.
Lund, 1918.
Greek Popular Religion. New York, 1940.
Griechische Feste von religioser Bedeutung mit Aussehluss der attischen. Leipzig,
1906.
History of Greek Religion. Oxford, 1925.
Homer and Myccnce. London, 1933.
Minoan-Mycenean Religion. Lund, 1927.
Myccntean Origin of Greek Mythology. London, 1932.
‘The New Inscription of the Salaminioi.' AJP 59. 385.
Primitive Time Reckoning. Lund/Oxford, 1920.
‘Sonnenkalender und Sonnenreligion.’ ARW 30. 14 1.
Studia de Dionysiis Atticis. Lund, 1900.
Odgers, M. M. The Latin Parens. Philadelphia, 1928.
Ogden, C. K., and Richards, I. A. The Meaning of Meaning. London, 1927.
Onians, R. B. ‘On the Knees of the Gods.' CR 38. 2.
Orde Browne, G. S. The Vanishing Tribes of Kenya. London, 1925.
Orwin, C. S. and C. S. The Open Fields. Oxford, 1938.
Oxford Book of English Verse. 1918.
Paget, R. A. S. ‘L’^ volution du langage.' Psychologic du langage, 92.
Human Speech. London, 1930.
Parry, M. Lesformules et la mltrique d'Homere. Paris, 1928.
‘Studies in the Epic Technique of Verse-making.’ HSC 41. 73.
‘Traces of the Digamma in Ionic and Lesbian Greek.' L 10. 130.
Paton, W. R., and Hicks, E. L. Inscriptions of Cos. Oxford, 1891.
Paul, H. Grundriss der germanischen Mythologie. Strassburg, 1900.
Pauly, A., and Wissowa, G. Realencyclopadie der klassischen Altertumswissen-
schaft. Stuttgart, 1894-1937.
Payne, H. G. G. ‘Archeology in Greece, 1934-5.' JHS 55. 147.
Perachora. Oxford, 1940.
PEARSON, L. Early Ionian Historians. Oxford, 1929.
Peel, R. F. ‘Rock-paintings from the Libyan Desert.' An 13. 389.
Pendlebury, J. D. S. Archeology of Crete. London, 1939.
PERERA, A. A. ‘Glimpses of Singhalese Social Life.’ IA 33. 143.
PERSSON, A. W. ‘Der Ursprung der elcusinischen Mysterien.’ ARW 21. 30 1.'
BIBLIOGRAPHY
593
Petersen, W. Quest fonts de historia gentium Atticarum. Schleswig, 1880.
Petrie, w. M. F. History of Egypt. London, 1894-1901,
Social Lift in Ancient Egypt. London, 1923.
Picard, C. Ephlsc et Claros. Paris, 1922.
'Sur la patrie et les peregrinations de Demeter,' REG 40. 3 36.
Pickard-CaMBRID GE, A. C. Dithyramb, Tragedy and Comedy. Oxford, 1927.
Playfair, A. The Caros. London, igog.
PLEKHANOV, G. V. Role of the Individual in History. London, 1940.
POLITES, N. G. ‘EisXoyal < 5 m 6 tit TpccyoOSia toO {XA^wikoO ActoO. Athens, I932.
‘On the Breaking of Vessels as a Funeral Rite in Modem Greece.'
JAI 23. 29.
TTotpaScbtJEis toO IAAtivikoO AaoO. Athens, 1904.
TTapoiulat. Athens, 1899-1902.
Prentice, W. K. ‘The Achsans.’ AJA 33. 206.
PRESCOTT, W. H. Conquest of Peru. (Everyman ed.) London, 1908.
PRITCHARD, J. B. Palestinian Figurines. New Haven, Conn., 1943.
Quain, B. H. ‘The Iroquois.' Meade 240.
Quiggin, E. C. Essays and Studies Presented to William Ridgeway. Cambridge,
19I3 ‘
Radcuffe-Brown, A. R. The Andaman Islanders. Cambridge, 1933.
‘The Social Organisation of the Australian Tribes.' O 1. 34, 206, 322,
426.
‘Three Tribes of Western Australia.' JAI 43. 143.
'Totemism in Eastern Australia.' JAI 59. 399.
Radin, P. Indians of South America. New York, 1942.
RADLOV, V. V. Aus Sibirien. Leipzig, 1884.
OBpastfhi napodHoU numepamypu mypucKux tinmen. Proben der
Volhslitteratur der tttrkischn StSmmc tend der Dstengarischen Steppe. Sc.
Petersburg, 1866-96.
Ramsay, W. M. Asianic Elements in Greek Civilisation. London, 1927.
Rattray, R. S. Ashanti. Oxford, 1923.
Three Tribes of the Ashanti Hinterland. Oxford, 1932.
Reichelt, K. ‘Der homerischen Genitiv auf -oio und verwandtes.’ ZVS 43.
68 .
Reinach, S. Orpheus. Paris, 1909. London, 1931.
Ripertoire des vases peints grecs et Itrusques. Paris, 1899-1900.
Traitl d' Ipigraphie gtecque. Paris, 1885.
Reitzenstein, R. Epigramm und Skolfon. Giessen, 1893.
Revillout, E. L'ancienne Egypte. Paris, 1909.
Ridgeway, W. The Early Age of Grace. Cambridge, 1901-31.
‘Measures and Weights.’ A Companion to Greek Studies. Cambridge, 1905.
'The Homeric Land System.’ JHS 6. 319.
Ritschl, F. W. Die alexandrinischen Biblfotheken und die Sammlung der homerischen
Gedichte dutch Pisistratus. Breslau, 1838.
Corollarium disputatfonis de bibliothecis Alexandrinis. Bonn, 1840-1.
Oo
594 STUDIES IN ANCIENT GREEK SOCIETY
Rivers, W. H. R. History of Melanesian Society. Cambridge, 1914.
Kinship and Social Organisation. London, 1932.
Tie Todas. London, 1906.
Robert, C. Aus der Anomia: archdologische Beitragen C. Robert dargebracht.
Berlin, 1890.
Bild md Lied. Berlin, 1881.
'Sosipolis in Olympiad MDA 18. 37.
Robert, L. Etudes anatoliennes. Saris, 1937.
Robertson Smith, W. Kinship and Marriage in Early Arabia, 2 ed. London, 1903,
Religion of the Semites. 3 ed. London, 1927*
Rodd, R. Homer’s Ithaca. London, 1927.
Rojas, A. V. ‘Kinship and Nagualism in a Tzeltal Community.’ AA 49. 578.
Roscher, W. H. Ausfiihrliches Lexikon der grieehischen und romischen Mythologie.
Leipzig, 1884-1937.
Selene und Verwandtes. Leipzig, 1890.
Roscoe, J. The Baganda. London, 1911.
The Bany ankle. Cambridge, 1923.
The Bakitara or Banyoro. Cambridge, 1 923.
The Bagesu and Other Tribes of the Uganda Protectorate. Cambridge, 1 924.
The Northern Bantu. Cambridge, 1915.
ROSSBACH, A. Untersucbungen iiber die romische Ehe. Stuttgart, 1853.
Rostovtzeff, M. History of the Ancient World. Oxford, 1927.
Roth, H. L. The Great Benin. Halifax, 1903.
Roth, W. E. Ethnological Studies among North-West Queensland Aborigines.
Brisbane/London, 1897.
ROUSE, W. H. D. Greek Votive Offerings. Cambridge, 1902.
Russell, R. V., AND Lal, R. B. H. Tribes and Castes of the Central Provinces of
India. London, 19x6.
Schachermeyr, F. Hethiter und Achaer. Leipzig, 1935.
Schaeffer, C. F. A. Cuneiform Texts of Ras Shamra. London, 1939.
Schapera, I. The Bantu-speaking Tribes of South Africa. London, 1937.
The Kloisan Peoples of South Africa . London, 1930.
Schefold, K. ‘Kleisthenes.’ MH 3. 59.
Schmidt, J. 'Reisefriichte.' MDA 5. 1x5.
Schoolcraft, H. J. Indian Tribes of the United States. Philadelphia, 1853-6.
Schuhl, P. M. La formation de la pensee grecque. Paris, 1934.
SCHWYZER, E. Griechische Grammatik. Munich, 1939.
Scott, J. A. 'The Relative Antiquity of Homeric Books.’ CP 14. 136.
'The Relative Antiquity of the Iliad and Odyssey.’ CP 6. 156.
The Unity of Homer. Berkeley, 1921.
SEEBOHM, F. The English Village Community. 4 ed. Cambridge, 1926.
Tribal Custom in Anglo-Saxon Law. London, 1911.
SEEBOHM, H. E. The Structure of Greek Tribal Society. London, 1895.
SEUGMAN, C. G. The Melanesians of British New Guinea. Cambridge, 1910.
Pagan Tribes of the Nilotic Sudan. London, 1932.
The Veddas. Cambridge, 1911.
BIBLIOGRAPHY
595
Seltman, C. T. Athens, Its History and Coinage. Cambridge, 1924.
Semple, E. C. Geography of tie Mediterranean Region. London, 1932.
Seyrig, H. 'Antiquites syriennes.’ S 20. 177.
Sheppard, J. T. 'Similes in the Iliad.’ JHS 32. 220.
Shewan, A. Lay ofDolon. London, 1911.
Shirokagaroef, S. M. Social Organisation of the Manchus. Shanghai, 1924.
Sigwart, G. 'Zur etruskischen Sprache.' Gi 8. 139.
Skene, W. F. Celtic Scotland. 2 ed. Edinburgh, 1908.
Smith, E. A. ‘Myths of the Iroquois.’ ARB 2. 77.
Smith, E. W., AND Dale, M. The Ila-speaking Peoples of Northern Rhodesia.
London, 1920.
Smith, S. Babylonian Legends of the Creation. London, 1931.
Smyth, H. W. Greek Melic Poets. London, 1906.
Smyth, R. B. The Aborigines of Victoria. London, 1878.
Spencer, B., and Gillen, F. H. Across Australia. London, 1912.
The Arunta. London, 1927.
Native Tribes of Central Australia. London, 1899.
— — Native Tribes of the Northern Territory of Central Australia. London, 1904.
Northern Tribes of Central Australia. London, 1899.
Stalin, F. Das hellenische Thessalien. Stuttgart, 1924.
Stalin, J. V. Dialectical and Historical Materialism. London, n.d.
Stawell, F. M. Homer and the Odyssey. London, 1909.
STERN, B. J. Lewis Henry Morgan. Chicago, 1931.
Strauss, C. L. ‘The Social Use of Kinship Terms among Brazilian Indians.’
AA 45. 308.
Sturtevant, E. H. Hittite Glossary. 2 ed. Philadelphia, 1936.
Talbot, D. A. Woman's Mysteries of a Primitive People. London, 1915.
Talbot, P. A. The Peoples of Southern Nigeria. London, 1926.
Tapun, G. 'The Narrinyeri.’ Woods NTSA.
Tarn, W. W. The Greeks in Bactria and India. Cambridge, 1938.
Thiel, J. H. ‘Zum vorgriechischen Mutterrecht.' K 24. 383,
Thomas, N. W. Kinship Organisations and Group Marriage in Australia.
Cambridge, 1906.
Thompson, J. E. Archeology of South America. Chicago, 1936.
Thomsen, V. 'Sur la parent^ de la langue etrusque.' OKD 1899. 1.
Thomson, B. H. ‘Concubitancy in the Classificatory System of Relation-
ship.' JAI 24. 371.
Thomson, G. Hschylus and Athens. 2 ed. London, 1946.
JEsckylus, Oresteia. Cambridge, 1938.
AEschyltts, Prometheus Bound. Cambridge, 1932.
‘The Greek” Calendar.’ JHS 63. 52.
Greek Lyric Metre, Cambridge, 1929.
Marxism and Poetry. London, 1946.
Thomson, J. A. K. Studies in the Odyssey, Oxford, 1914.
Thuenwald, R. Economics in Primitive Communities. London, 1932.
596 STUDIES IN ANCIENT GREEK SOCIETY
Tiumenev, A. J. « Pa 3 floztceHue podoeoso cmpon u covpunnbHan peeo/voifux
VI I -VI eg. 8 rpei ( uu . t > Marr and Prijuzhin 32.
Tod, M. N. Greek Historical Inscriptions. Oxford, 1933.
TOEPFFER, J. Attische Genealogie. Berlin, 1889.
‘Theseus und Peirithous.' C. Robert AA 30.
Torday, E., and Joyce, T. A. ‘Ethnography of the Bahuana.’ JAI 36. 272.
Toutain, J. L’ economic antique. Paris, 1927. Economic Life of the Ancient
World. London, 1930.
Tritsch, F. ‘Die Agora von Elis und die altgriechische Agora.' JOA 27. 63.
‘Die Stadtbildung des Altertums und die griechische Polis.' K 22. 1.
Tylor, E. B. Primitive Culture. 3 ed. London, 1891.
UsENER, H. Altgriecliseher Vershau. Bonn, 1887.
Valmin, M. N. Swedish Messenia Expedition. Lund, 1938.
Vambery, A. Travels in Central Asia. London, 1864.
Van GenNEP, A. L’etat actuel du probllme totfmiquc. Paris, 1920.
— — Les rites de passage. Paris, 1909.
Vendryes, J. ‘La position linguistique du celtique.’ PBA 33.
VlCO, G. Principi di scien^a nuova. Naples, 1744.
VlNOGRADOFF, P. Growth of the Manor. London, 1905.
Outlines of Historical Jurisprudence. London, 1920-2.
Wace, A. J* B. ‘Chamber Tombs of Mycenae.’ Arc 82.
‘Excavations at Mycenae.’ ABS 25. 3.
‘Mycenae.’ An 10. 405.
‘Mycenae, 1939.’ JHS 59. 210.
‘The Treasury of Atreus.’ An 14. 233.
-and Thompson, M. S. Prehistoric Thessaly. London, 1912.
Wackernagel, J. ‘Sprachliche Untersuchungen zu Homer.' Gl 7. 161.
Wade-Gery, H. T. ‘Kynaithos.’ Greek Poetry and Life (Oxford 1936), 56.
Wainwright, G. A. The Sky-Religion in Egypt. Cambridge, 1938.
Walde, A., and Pokorny, J. Vergleichendes Wortcrbuch der indogtrmanischen
Sprachen. 2 ed. Leipzig, 1928-33.
Walpole, R. Travels in Various Countries of the East. London, 1820.
WARRINER, D. Land and Poverty in the Middle East. London, 1948.
Webster, H. Primitive Secret Societies. 2 ed. New York, 1932.
Weniger, L. 'Das Hochfest des Zeus in Olympia.’ K 5. 1.
WerthaM, F. Dark Legend: a Study in Murder. New York, 1941.
. Wesendonk, O. G. VON. ‘Zur Verwandtenehe bei den Arsakiden.' ARW
30. 383.
Westermann, A. Scriplores rerum mirabilium Graeci. Brunswick, 1839.
WESTERMARCK, H. H. History of Human Marriage. London, 1901.
The Origin and Development of Moral Ideas. London, 1 906-8.
Westlake, H. D. Thessaly in the Fourth Century B.c. London, 1935.
WESTRUP, C. W. 'Recherches sur les formes antiques de mariage dans
l’ancicn droit romain.’ OKD (1899) 30. 1.
BIBLIOGRAPHY - 597
Wilamowitz-Moellendorpf, U. von. Der Glaube der Hellehen. Berlin,
1931.
Griechische Verskunst. Berlin, 19-21.
Die litas und Homer. Berlin, 1916.
XJeber die ionische Wanderung. (Sitzberichte der preussischen Akademie
der Wissenschaften.) Berlin, 1906.
Wilhelm. A. ‘Die lokrische Madcheninschrift.’ JOA 14. 163.
Williamson, R. W. Social and Political Systems of Central Polynesia. Cam-
bridge, 1924.
Wittfogel, K. A. Wirtschaft und Gcsellschaft Chinas. Leipzig, 1931.
Wolf, F. A. Prolegomena ad Homcrtim. Halle, 1795.
Wollaston, A. F. R. Pygmies and Papuans. London, 1912.
Wood! R. Essay on tie Original Genius of Homer. London, 1769.
Woods, J. D. Native Tribes of Sottth Australia. Adelaide, 1879.
Woolley, L. ‘Tal Atchana.’ JHS 56. 125.
Wycherley, R. E. ‘The Ionian Agora.’ JHS 62. 22.
Xanthoudides, S. Vaulted Tombs of Mesara. Liverpool, 1924.
Yeats, W. B. Essays. London, 1924.
Zanotti-Bianco, U. ‘Archaeological Discoveries in Sicily and Magna
Graecia.' JHS 57. 238.
Zuckermann, S. 'The Biological Background of Human Social Behaviour.’
(Proceedings of the Second Conference on the Social Sciences at the Institute of
Sociology .) London, 1936.
PERIODICALS
An Antiquity. London, 1926-.
Ate Archeolagia . London, 1770-.
As Anthropos. Vienna, 1905-.
AA American Anthropologist. Washington/New York, 1888-.
ABS Annual of the British School at Athens. London, 1894-.
AJA American Journal of Archeology. Concord, 1897-.
AJP American Journal of Philology. Baltimore, 1879-.
ARB Annual Reports of the Bureau of Ethnology. Washington, 1881-.
ARW Archivfiir Rcligionswissenschajt. Freiburg, 1898-.
AS Annie sociologique. Paris, 1 896-.
BCH Bullltin de correspondance hellenique. Paris, 1877-.
C Caucasica. Leipzig, 1924-.
CP Classical Philology. Chicago, 1906-.
CQ Classical Quarterly. London, 1906-.
CR Classical Review. London, 1887-.
Ge Georgica. London, 1935—.
Gl Glotta. Gottingen, 1907-.
HJA Harvard Journal of Asiatic Studies. Cambridge, Mass., 1936-.
HS Harvard Studies in Classical Philology. Cambridge, Mass., 1890-.
IA Indian Antiquary. Bombay, 1872-.
IF Indogermanische Forschungen. Strassburg, 1891-.
JAI Journal of the Anthropological Institute. London, 1872-.
JAO Journal of the American Oriental Society. Baltimore, i88o~.
JHS Journal of Hellenic Studies. London, 1880-.
JOA Jahresheft des osterreichischen arch'dologischen Institutes. Vienna, 1897-.
JOG Journal of Obstetrics and Gynecology. London, 1902-.
JP Journal de psychologic. Paris, 1908-.
JPS Journal of the Polynesian Society. Wellington, 1892-.
JRS Journal of Roman Studies. London, 1911-.
K Klio. Leipzig, 1901-.
KF Klcinasiatischc Forschungen. Weimar, 1930-.
L Language. Philadelphia, 1 924-.
M Man. London, 1901-.
MDA Mitteilungen des deutsches archaologisches Instituts: Athenische Abteilung
Berlin, 1876-.
MH Museum Helveticum. Basel, 1943-.
MI Man in India. Hyderabad, 1920-.
NJK Netie Jahrhucher fiir das klassisches Altertum. Leipzig, 1898-.
NRH Nouvelle revue historique du droit franyais et Granger. Paris, 1855.
O Oceania. Melbourne/London, 193 1-.
OKD Oversigt over det hongelige Danshe Videnskabemes Selskabs Forhandlinger
Copenhagen, 1816-.
PERIODICALS
599
OL O rientalisclc Literaturzeitung . Leipzig, 1897-.
PBA Proceedings of the British Academy. London.
PRS Proceedings of the Royal Society. London, 1903-.
REG l&vuf des Itudes grecs. Paris, 1887-.
RHA Revue hittite et asianique. Paris, 1930-.
RM Rheinisches Museum fiir Philologie. Frankfurt-am-Main, 1851-.
RP Revue de philologie. Paris, 1877-.
S Syria. Paris, 1919-.
SJA Southwestern Journal of Anthropology. Albuquerque, 1945-.
VDI BecmHUK dpeeneu ucmopuu. Moscow/Leningrad, 1947-.
YCS Yale Classical Studies. New Haven, 1928-.
ZA Zeitschrift fiir Assyriologie und verwandte Cehiete. Berlin/Leipzig, 1892-.
ZVF Zeitschrift fiir vergleichende Sprachforschung. Berlin, 1852-.
INDEX TO MAPS
The references are to Map XII except where otherwise indicated.
References to the other maps are given only for places not marked on
Map XH.
Abdera I Bb
Anaphe Fi
Abydos Ga
Andania Bg
Achaia, Peloponnese BCe
Andros EFe
Achaia Phthiotis Cc
Antandros Gb
Acharnai BCk
Apameia IBb, X Ad
Acheloos Bd
Aphidna Cj
Achilles, Harbour of Ci
Aphrodisias Ga
Adramyttion GHc
Apodotoi BCe
Aigai Ce, He
Arabia I Cc
Aigeira Ce
Araithyrea VII Ca
Aigialeos Bk
Arakynthos IX Cb
Aigilia Df
Araphen Ck
Aigilips IX Bb
Arcadia BCf
Aigina Df
Argos Cf
Aigion Ce
Armenia I Ca, X Dc
Aiolis GcHd
Arne Cc
Aitolia Bd
Artemision Dc
Akamania ABd
Asine Cg
Akragas I Ac
Askra Aj
Akraiphion Ai
Asopos, Achaia Ce
Akriai Ch
— , Bceotia Cj
Akte Ea
— , Laconia Ch
Alabanda Hf
Aspledon Ai
Alalkomenai Ai
Assos Gc
Alea Cf
Astakos Ae
Aliphera Bg
Astypalaia Gh
Alope Cd
Athens (Athenai) Df
Alpheios Bf
Athos Ea
Alyzia Ad
Atrax Cb
Amastris I Ca, X Be
Attica (Attike) De
Amisos X Cc
Aulis Bi
Amnisos Fk
Axios I Bb
Amorgos FGh
Azenia Df
Amphilochia Be
Amphipolis I Bb
Bithynia X Be
Amprakia Ac
Bceotia (Boiotia) CDe
Amyklai Cg
Boghaz-keui I Ca, X Cc
Anagyrous Df
Boiai Di
Anaia Hf
Boibe Cc
602 STUDIES IN ANCIENT GREEK SOCIETY
Boion Aa
Bospotos I Bb, X Ac
Bouprasion Bf
Bovianum I Ab
Branchidai Hg
Brauron Df
Byblos X De
Byzantium (Byzantion) I Bb, X Ac
Cappadocia I Cb, X Cc
Carchemish X Dd
Caria Hg
Caucasus (Kaukasos) I Ca, X CG
Chaironeia Ce
Chalkidike Da
Chalkis De
Chios FGe
Chryse Gb
Cilicia (Kilikia) X Ce, I Cb
Corinth (Korinthos) Cf
Corsica I Ab
Cyclades E-Ge-h
Cyprus I Cbc, X Ce
Eleusis De
Eleutherai Bj
Eleutherna Ek
Elis Bf
Enipeus Cc
Epeioi III Ab
Epeiros ABab
Ephesos Hf
Ephyra, Akamania Bd
— Corinth HI Bb
— Thesprotia Ac
Epidauros Df
Epidauros Limera Dh
Eretria De
Etymanthos Bf
Erythrai Ge
Euboia Is. Dd
EuboiaMt. VI
Euenos IX cb
Euphrates X Dc
Euripos Bi
Europos Cb
Eurotas Ch
Eurytanes Bd
Dalmatia I Ab
Danube I Ba, X Ab
Daskylion Hd
Dekeleia Cj
Delos Fg
Delphi (Delphoi) Ce '
Delta I Cc
Dikte Fk
Dimini C Dc
Dion Da
Dodona Ab
Doliones Ha
Dolopia Be
Dorion Bg
Doris Cd
Doulichion Be
Dyme Be
Egypt I Cc
' Elaia Hd
Elateia Cb
Eicon Bj
Fucinus L. I Ab
Gargara Gb
Gerenia Ch
Golgoi I Cb
Gonnos Cb
Gortyna Ek
Gournia Fk
Granikos R Ha
Gryneion Hd
Gyrton Cb
Gytheion Ch
Halai Araphenides Ck
Haliakmon Ca
Haliartos Aj
Halikarnassos Hg
Halos Cc
Halys I Ca, X Cc
Hebros I Bb
Helike Ce
Hellespont Ga
INDEX TO MAPS
Helos Ch
Heniochoi I Ca, X Db
Hephaistia Fb
Heraia Bf
Heraion Cf
Hermione Dg
Hermos Hd
Hierapytna Fk
Hyettos Ai
Hymcttos Ck
Hysiai Bj
Ialysos Hi
Ida, Crete Ek
— , Troad Gb
Ikaria, Attica Ck
— > Aegean G£
Ikos Dc
Ilissos Ck
Illyria I Aa
Imbros Fa
Inachos Cf
Iolkos Dc
Ionia GeHf
Ios Fh
Ioulis Ef
Isthmus of Corinth Df
Ithaca (Ithake) Ae
Ithome Bg
Itonos Cc
Joppa I Cc
Kadesh I Cb, X De
Kaikos He
Kalauria Dg
Kalydon IX Cb
Kalymnos Gh
Kameiros Hi
Kardamyle, Chios Gd
— , Messenia Ch
Karnos IX Bb
Karpathos Hj
Karthaia Ef
Karyai Cg
Karystos Ee
Kaukones HI Ac
Kaukoniatai I Ca, X Be
Kaunos I Bb, X Ae
Kaystros He
Kelenderis, Argolis Df
— Cilicia I Cb
Kenchreai VI
Keos Ef
Kephallenia Ae
Kephisos, Attica Bk
— Bceotia Cd
Kerkyra I Bb
KillaGb
Kimolos Eh
Kissa X Dg
Kissos HE Ba
Kissos Mt. Da
Kithairon Aj
Klaros He
Klazomenai Ge
Kleonai Cf
Knidos Hh
Knossos Fk
Kolchis I Ca, X Db
Kolophon He
Komana I Ca, X Cc
Kopais L. Ai
Korax IX Ac
Korone, Messenia Bh
— , Thessaly Dc
Koroneia Aj
Kos GHh
Kourion I Cc
Krannon Cc
Krisa Ce
Krokyleia IX Ab
Kroton I Ab
Kydonia Ek
Kyllene Be
Kyllene Mt. Cf
Kyme Gd
Kynouria Cg
Kyparissia Ch -
Kythera Di
Kythnos Eg
Kyzikos Ha
604 studies IN ANCIENT GREEK society-
Lykone VI
Labranda Hg
Laconia (Lakedaimon) Ch
Ladon Bf
Lamia Cd
Lampsakos Ga
Laodikeia, Cappadocia I Ca, X Cc
— Syria I Cb
Lapathas I Cb
Larisa, Argos III Be
— , Elis III Ab
— , Lydia He
— , Thessaly Cb
— i Troad Gb
Larisa Kremaste Cd
Larisaiai Petrai HI Db
Larisos Be
Larymna Bi
Las Ch
Latium I Ab
Latmos Hf
Lato Fk
Lebadeia Ai
Lebedos Ge
Lechaion Cf
Lelantos Bi
Lemnos Fb
Lepreos Bg
Lema Cf
Lesbos FGc
Letrinoi Bf
Leukadia IX ABa
Leukas Ad
Lenkosyria I Ca, X Cc
Leuktra, Boeotia Aj
— , Laconia Ch
Libya I BCc
Lilybaion I Ab
Limnai Cg
Lindos Hi
Liparai Is. I Ab
Lokris Epiknemidia Cd
Lokris Opountia De
Lokris Ozolis BCe
Lokroi Epizcphyrioi I Ab
Lucania I Ab
Lycia (Lykia) I Cb, X Be
Lydia Hde
Macedonia BCDa
Magnesia on Maiandros Hf
— under Sipylos Hd
— , Thessaly CbDc
Maiandros Hf
Makisros Bg
Malea Di
Malis Cd
Mantineia Cf
Marathon De
Maroneia I Bb
Mases Dg
Megalopolis Cg
Megara Df
Meiiboia Cb
Melos Eh
Mesopotamia I Cb
Messara Fk
Messe Ch
Messenia (Messene) Bg
Messogis Hef
Metapontion I Ab
Methymna Gc
Midea Cf
Milatos Fk
Miletos Hf
Mochlos Fk
Molottoi Abe
Mossynoi X Cc
Mounychia Bk
Mycenae (Mykenai) Cf
Mykale Hf
Mykalessos Bj
Mykonos Fg
Mylasa Hg
Myous Hf
Myrine, Ionia Hd
— , Lemnos Fb
Myrrhinous Df
jhlysia He
Mytilene Gc
Nasamones I Be
Nauplia Cf
Naxos Fg
INDEX TO MAPS
605
Neda Bg
Ncleia IV
Nemea VI
Neonteichos Hd
Neritos IX Ab
Nikaia I Bb, X Ad
Nisaia Bk
Nisyros Hb
Notion Hf
Oche, Ee
Oitylos Ch
Olbe I Cb, X Cd
Olenos Be
Olizon Dc
Oloosson Cb
Olympia Bf
Olympos Ca
Onchestos Aj
Ophioneis Bd
Opous De
Orchomenos Ce
Ormenion Cc
Orneai Cf
Orontes X Cd
Oropos, Bceotia De
— , Thesprotia Ac
Ossa Cb
Othrys Cc
Pagasai, Gulf of CDc
Paionidai Ck
Palaikastro Gk
Pale IX Ac
Pallene Da
Pamisos.Messenia Bg
— , Thessaly BCc
Pamphylia I Cb, X Be
Paphlagonia I Ca, X Be
Paphos I Cb, X Be
Parnassos Ce
Pamon Cg
Paros Fg
Patara X Be
Patmos Gg
Patrai Be
Pedasa Hg
Pedasos, Messenia III Be
— , Troad, HI Da
Peiraieus Df
Pelasgiotis Cb
Pelion Dc
Pellene Ce
Peneios, Elis Bf
— , Thessaly ’h'C.’oc
Peparethos Dc
Peraia Ce
Pergamos He
Petra Ca
Phaistos Ek
Phaleron Bk
Phanagoreia I Ca, X Ca
Pharai, Achaia Bf
— , Laconia BCh
Pharygai Cd
Phasis R. X Db
Pheneos Cf
Pherai Cc
Phigalia Bg
Phleious Cf
Phokaia Gd
Phokis Cd
Pholegandros EFh
Pholoe Bf
Phrygia GHa
Phylake, Thesprotia Ac
— , Thessaly Cc
Picenum I Ab
Pindos Be
Pisa Bf
Pitane Gd
Plakia Hd
Plataia Aj
Pleuron Be
Poteidaia Da
Potniai Bj
Praisos Gk
Prasiai Cg
Priene Hf
Probalinthos Ck
Pronnoi Ae
Propontis Ha
Psamathous Ci
Psophis Bf
606 STUDIES IN ANCIENT GREEK SOCIETY
Psyra Fd
Pteleon Dc
Pygela Hf
Pylos, Messenia Bh
— , Triphylia Bg
Pyrasos Cc
RachmaNI Cb
Ras Shamra X De
Rhamnous Cj
Rheneia Fg
Rhodes (Rhodos) Hi
Rhypes Be
Rome I Ab
Salamis, Attica Df
— , Cyprus I Cb, X Cc
Salmone Bf
Samos, yEgean Gf
— , Kephallenia Ae
Samothraike Fa
Sangarios I Cab, X Bd
Sardes He
Sardinia I Ab
Schoinos Bi
Scythia I Ba, X Aa
Seriphos Eg
Sesklo IV
Sestos Ga
Sidon I Cc, X De
Sigeion Gb
Sikinos Fh
Sikyon Cf
Sinope X Cc
Siphnos Eg
Sipylos He
Sithonia Da
Skamandros Gb
Skepsis Gb
Skiathos Dc
Skione Db
Skyros Ed
Smyrna He
Soloi I Cb, X Cd
Sounion Df
Sparta Cg
Spercheios Cd
Sporades Ggh
Strymon I Bb
Sybaris I Ab
Syme Hh
Syracuse (Syrakousai) I Ac
Syria I Cb
Syros Eg
Sythas VI
TAINARON Ci
Tanagra De
Taphiai Is, Ad
Taphos IX ABc
Taras I Ab
Tauris I Ca, X Bb
Tauros I Cb, X Cd
Taygetos Cg
Tegea Cg
Telos Hi
Tempe Cb
Tenedos FGb
Tenos Ff
Teos Ge
Teuthrania He
Teuthrone Ch
Thalamai Ch
Thasos Ea
Thebai, Thessaly Cc
Thebes (Thebai) De
Thelpousa Bf
Themiskyra I Ca, X Cc
Thera Fi
Thespiai Aj
Thesprotia Ac
Thestios IX Ca
Thisbe Aj
Thouria Bg
Thrace (Thraike) I Ba
Thyateira Hd
Thyssos Ea
Tiryns Cf
Tmolos He
Torone Db
Trachis Cd
Tralles Hf
Trikka Bb
Trikolonoi Cg
INDEX T O MAPS
607
Triphylia B£
Tritaia Bf
Tyre I Cc, X De
Troad (Troas) Gb
UGARIT X De
Troizen D£
Troy (Troia) Gb
Xanthos X Be
Tsangli IV
Zakynthos Af
Tsani Magoula IV
Zerelia IV
GENERAL INDEX
Abas, Abantes 285, 382, 391
Ablution 223-6
Achajans 385-4x6, 430-2; and
Kassitcs 298; conquered Knossos
373; descent from Zeus 286; dia-
lect 5x7-8, 523-4; piracy 322,
370. '
Achaia 385-6, 430
Achaios 386
Achillas: ancestry 378-9, 397-8,
400; his horse 345; as minstrel
497
Adcock 299, 327
Adonis 498-9, 513
Adoption 89, 94-6, no, 333, 4X9
Adultery 98, 140, 143
Asneas 328
TEolian migration 396, 518, 542-4
>Eolic dialect 324, 399-400, 407,
515-26
African monarchies 156-9
Agamemnon: ancestry 408; domain
393-4; king of Kyme 544; tomb
410.
Agamemnoneion Genos 408
Age-grades 45, 61, 64
A&Ra 145. 4 8 3 » 49 8
Agenor 375-7
Aglaia 340
Agori 354, 362
Agorios 405, 409
Agriania 195-6, 226-7, 485, 499
Agriculture 22-3, 33; migratory,
34, 88, 91, 349; women in, 42,
150,204.
Aiakidai 387-90
Aias 259: see Ajax
Aigina270, 280, 388-90, 397
Aiolos 391, 417
jifs«fo>s495
Aitoloi 353, 409-10, 524
Ajax X2i, 345, 389
Pp
Akamania 360, 421, 425-7
Akrisios -382
AkrSpolis 257, 362
Aktaion 224
Alalkomenai 129
Alcuin 577
Alesiai 129
Algonkins 91
Alitherses 427
Alkaios 465-74, 496
Alkinoos 360-1, 4x8-20
Alkmaion 342
Alkman 219, 465-74, 492, 498,
560
Alkmeonidai 192, 198
Allen 515, 569-70
Alphabet 529, 575
Alpheios 223-4
Althaia 432
Althaimenes 324
Alyzeus 425
Amaltheia 115, 250, 256, 344
Amazons 180-3, 287, 290, 407, 429
Amber 293
Amnisos 231, 245, 253, 283
Amphiaraos 330
Amphion 402
Amphipolis 324
Amyklai 410
Anaklethra 13 1, 233
Ancestor worship 39-40, 44, 499
Andania 125, 254
Andromeda 382
Anglo-Saxons 82, 308, 314, 326,
496, 546, 576-7
Anigros 226
Ansted 312-3
Apatouria 145, 544
Apheidantes 132
Aphrodite: leader of Charites 341;
goddess of childbirth 219; in
Cyprus 165, 413.
GENERAL INDEX
6lO
Apotkfa 323
Apollo: birth 276; Delian 551, 563;
Delphic 1 1 5, 216, 227, 346;
Dorian 102; Hekatos 230; Klarios
'270-1, 286; god of Lapithai 197;
as prophet and musician 215, 333,
484-5; wolf 1 14, 163.
Apollonios 195
Apsu 247, 263
Arabs 25, 305, 326, 336
Arai 136
Araithyree 129
Archandrou Polis 386, 401
Archilochos 474
Arclon 1 14, 359, 363-4
Ares ’286, 4x8
Arete 418-20
Argonauts 175, 195, 546
Argos, name of 171
Ariadne 254
Arion 565
Aristeides 257
Aristotle: on Athenian democracy
125, 363-4; Hellenes 292, 398;
Homer 561-2, 578; pilis 138,
351; Politics 143; tetrameter 476;
tragedy 467; tribal system 105-7;
trimeter 474.
Arkas 131-2,276
Armeni 25, 183, 232
Arne 397
Arrhephoria 222
Arrhephoroi 222, 234, 247
Arrian 162, 182
Arsacids 162
Artemis 269-80; Alpheaia 223;
Angelos 209; at Brauron 483; at
childbirth 219, 244; Delphic 294;
Ephesian 182, 282, 482; Hekate
229-30; Lousia 226; • Lygodesma
218, 246; leader of Moirai 338,
340; Mounychia 279, 406; Orthia
429; at Tauris 408; as waterfowl
429, 506.
Artemisia 167-8
Arunta 39, 66, 74, 86, 352
Ashurbanipal 509
Asklepiadai 333
Asklepios2i6, 246, 333, 395
Assyria 25, 380
Ate 1 37
Athena 257-68; Alea 134; birth
244; Itonia 397; Polias 108, 126,
186, 222, 482; snake goddess 116,
124; Trojan 278; Victory 219.
Atossa 162
Atreus 382-3, 404, 407-9
Atropos 334-5
Aurignacian 52
Autesion 402
Auxesia 221
Axe, double 251, 290
Axios X84, 400
Azanes 13 1
Aztecs 91, 100, 150
Babylonia 23-8, 161, 239, 298
Bacchants 459, 485
Bachofen 41, 144, 175
Baden-Powell 304
Baganda 156
Bakchidai 201-2, 565
Bakchylides 474
Ball dances 214, 234
Ballad measure 450, 464, 470
Bantus 36, 52, 120, 159, 207-9,
336
Baptism 46, 95
Barter 33, 356
Bathonga 120, 303
Bede 307
Bellerophon 164-5, 177, 190-1,
328, 382
Belos 375, 379
Benin 156
Bergk 474
Bias 191
Binary form 450-1
Blindness of minstrels 550
Bloch 307
Boiotoi 259, 263, 390, 395-400
Book of the Dead 1 19
Borough English 153
GENERAL INDEX
Bosporos 284
Bouphonia 122
Boutcs, Boutadai 107-8, 120-1,
126, 1S6, 265-6, 326, 392
Bowra 474, 532-3
Brea 323, 325, 328
Brfal 211
BrifTntilt 41, 70, 144, 157, 204-5,
214, 462
Britomartis 255. 275
Bronze 22-3, 28, 355
Bucher 446-7, 462
Buck 309
Burial, contracted 48, 55, 210;
early Greek 503-5: in jars 249-
50; Minoan 249
Burn 409
But)- 370
Byblos 26
Byzantine land-tenure 203, 312,
31S
Caedmon 459
Caesar 92, 527
Calendar 105-6, 210-1, 236, 269,
569. 482
Camp, tribal 21, 352
Cantnbri 141
Capstan culture 56
Carians 166-71; crests 290, 371:
double axe 290; Early Cydadic
and Hclladic 177, 193, 374; in
Ionia 169, 271, 542.
Caroline Is. 241
Carthaginians 322
Cat svorship 214
Cato 93
Cattle-raising 33, 42, 298
Cattle-worship 251
Caucasus 176, 179, 182, 261, 278,
298, 387
Caudwell 461
Cavaignac 407
Caves 53-5, 224, 231-2. 245-6,
Z49-50, 253
Celtic 80, 96, 136, 212, 344
6ll
Chadwick 412, 462-3, 531, 576
Chalkis 315, 466
Chariot 298, 407
Charitcs 339-41, 483-5
Chemmis 379, 381
Childbirth 204-10, 218-20, 230,
244-6, 253, 339
Childc 24, 34-5, 52
China 36, 46, 99, 141, 155-6, 314
Chios: dialect 515, 525, 544; centre
of Homeridai 490, 549.
Cicero 215, 571
Circumcision 47-8
Clan cults 123-32, 479-81
Clothes 336, 486
Clyteinnestra 200, 407, 429-30
Commodities 556, 358
Communism 319-21, 331, 339
Cook 179, 287, 291
Copper 22, 26, 28, 355
Corinth, epic recitals at 565.
Cornford 232-4, 237
Cornucopia 344
Coronation 46
Council 46, 360, 363
Craft clans 332
Cremation 504-5
Crete 26-9: dialect 399.
Croatian minstrelsy 529
Cureau 45
Cyclades 28-9, 177. * 75 * 375 * 4*7
Cyclic poems 554-6, 562-3
Cyprus 26, 510-4; Achaan settle-
ments 386-7, 399, 401; dialect
399 * 5 * 7 -
Cyrus 336
Daf.hu 128
Dahomey 157
Dnidalos 285, 333
Dafmon 337—8
Daft 330, 338
Dakotas 76
Dalmatians 320
Damia 221
Danac 383
GENERAL INDEX
6l2
Danaos X28, 132, 377-8 5, 468
Dante 527
Danubian culture 34, 238, 240
Datdaneis 225
Dareios 162
Darwin 70, 84-5
DasmSs 326, 329-30, 338
Deliades 485
Delos, feast of Apollo at 490-5,
545 . 551
Delphi 102-3, 1x5, 216, 294, 342,
376,389,485
Delphic Oracle 328
Deme 109, 112-3, 3x3-4, 326-7,
351, 361
Demeter 249-5 6 ; Achaia 124, 131-
2; mother of Artemis 230;
Chamyne 289; Dorian 102; Eleu-
sinian 132, 231-7. 545 ; Erinys
• 342; partner of Herakles 288-9;
of Iasion 256, 292; at Orcho-
menos 193, 3755 Pelasgis 377;
Thesmophoros 13 1, 221-3.
Dtmiourgol 355, 359, 363-4
Demodokos 459, 485, 494, 581
Demos: see Deme
Dendra 504
De Pradenne 36, 53
Despoina 220
Deubner 222
Dikaiarchos 137-8
Dike 135, 339, 345
Dimini 184, 197, 238, 375
Dio Chrysostom 289
Diodoros 180-2, 320-1
Dionysus: Agrianios 195-6, 226,
499; Auxites 221; Corinthian
246; infant 224; Kittos 196;
Melpomenos 122.
Dirge 481-2, 484-5
Division of labour 22-3, 39, 42-5,
354 - 5 * 358
Dobu 61
Dodona 171, 292, 397-8, 400
Dolon, Lay of 573—5
Dorians: conquest of Crete 342; of
Laconia 273, 316, 409; in Homer
416, 501; tribal system 102, 166;
use of iron 29.
Doric dialect: 137. 254, 318, 351,
399, 466, 524
Doros 391
Doplichion 423
Dracontius 5x0
Dravidian 69
Dysaules 129, 131
Early Cycladic 374
Early Helladic 374, 377
Earthy 48, 218, 225, 241
Echelas 408, 543
Egypt 22-6, 28
Egyptian kingship 23, 50, 159
Eileithyia 115, 244-5, 253, 275,
292, 335 . 483
Eirene 339
El 377
Elektra 402
Eleusinian Mysteries: see Demeter
Eleutherolakones 393
Elysian Fields 513
Enclosure Acts 300
Endymion 215
Engels 86, 93, 300, 322
Enipeus 191-2, 194
Epameinondas 121 N
Epaphos 284-5. 379
Epeioi 264, 399
Epeiros 115, 395
Ephesos 181-2, 269-76, 294, 544-5
Ephyra 190, 324
Epicharmos 230, 236
Epicurus 143
Epidauros 170, 216, 257, 275, 391
Epilepsy 215
Epode 466-8
Eratosthenes 369-70, 409-10, 542
Erechtheus 1 16, 120, 126, 266, 392
Eretria 124
Erichthonios 222, 262
Erinyes: 117, 136-7, 220, 226, 254,
333 . 34 i- 2 . 345 - 6 . 395, 468
Erosantheia 234
GENERAL INDEX
613
Eskimos 60
Esmein 300, 302, 320
Eteokretes 172, 250
Ethiopians 141
Etruscans 92, 98-9, 141, 173-5,
179, 290, 322-3
Euenos 427
Eumaios 421, 580
Eumelos 551
Eumolpos 127, 13 1, 362
Euneos, Euneidai 122, 175, 191,
196, 356
Eunomia 339
Eupatridai 363-5
Euphrates 22, 150
Euphorbos 502
Euphrosyne 339
Europa 123, 376-7, 379
Euiystheus 174, 382-3
Eurytion 388-9
Evans 144, 178, 238-9, 487, 529
Exogamy 35, 41-5, 56-9, 66-71,
95, 138-9,153
Exorcism 460
Familia 92-3, 109-10, 139
Family 72, 77, 83-5, 297, 314
Famell 228, 244, 284-5
Faroe Is. 450
Ferguson 144, 300
Fick 516, 528
Figurines 237-48, 271
Fiji 65-6
Firstfruits 51
Fison 67, 84
Food-gathering 21, 33, 42, 204
Forrer 40 x
Forsdyke 193
Fotuna 65
Frazer 36, 122-3, 158
Gardiner 115
Gardner 106-9, 112
Garos 155
Garstang 182, 407
Gauls 97, 527
Gavelkind 307, 326
Gela 126
Genesis 425
Genetyllides 245
Genius 337
Geometry 317
Gtomiroi 363—4
Gephyraioi 123-4
Giras 329-3, 345
Germanic 80, 82, 136, 204, 212,
307 . 527 , 531 . 535 - 546 , 576
Germans III, 135, 139, 308, 431,
527
Gesture 244, 445-6
Glaukos son of Hippolochos 165,
544
Glaukos son of Minos 119
Glaukou Demos 165
Glotz 144
Gnorlsmala 3 36-7
God, idea of 50, 1 58, 246, 414
Goethe 147, 323, 443, 456
Goldenweiser 57
Gortyna in, 139, 178, 342
Graikoi 398
Granet 36, 155
Gras 408, 543
Gronbech 135, 332
Grate 106, 109, 200
Groves, sacred 1x3
Guilds 332, 357
Guiraud 321-2
Gumelnita 238
Gurdon 151-4
Gyges 406
H addon 33
Hadow 454
Hagia Triada 485, 488
Hair-cutting 47-8, 213
Halikamassos 103, 167
Hall 161
Hardy 95
Harmodios 123-4
Harrison 1x7, 144, 232, 284
GENERAL INDEX
614
Harvest, Greek 233, 309-10
Hawkins 202
Hector 416, 481, 536-7
Hecuba 260, 406, 416, 481, 507
H(Arai 362
Hegemone 221
Hekate 229-31, 236
Helen 395, 415-6, 429-30, 481,
505-14
Helikon 104, 390, 459, 485
Helios 263, 324-5, 491
Hellanikos 171, 542
Hellas 171, 387, 395-6
Hellen 369-70, 391
Heniochoi 387
Hepa 180
Hephaistos 172, 178, 261, 286,
333,413-4, 513
Hera 280-92; Argive 246, 263,
340, 377, 413; cow-faced 240;
mother of Eileithyia 245, 287; in
herbal lore, 218-9.
Heraia 240
Herakleitos 345
Herakles: Argive 282, 287-8, 383;
Cretan 288; Dorian 102, 264,
392; Hera’s partner 287-8; and
Hippokoon 288, 428, 492; name
289; servant of Omphale 174; at
weddings 292,
Hercules 291
Hereros 157
Hermes 128, 172-3, 337, 339
Hero worship 117, 120
Herodotus: at Chemmis 379; as
ethnologist 140, 143; on Homer
508-9; as prose chronicler 564.
Hesiod: ancestry 566; creator of
Greek theogony 287; as minstrel
459, 478, 491, 496; name 566;
society at Thespiai 485.
Hestia 363
Heurdey 193, 427
Hicron 126-7
Hindus 111, 338, 360
Hipparchos 123, 568-71
Hippodameia 265, 393, 403-5
Hippokoon 288, 428, 492
Hippokrates 142, 227
Hippostratos 567
Hittites26, 179-80, 401-2; collapse
of empire 407; at Ephesos 182-3,
269-70; law 160; mother-goddess
235, 407.
Homer: birthplace 547-9; creator of
Greek theogony 287; as minstrel
490-1, 508, 551-2; name 550 -x.
Homeric Hymns 490-1, 554
Homeric minstrelsy 458, 478, 488-
92, 497
Homeric simile 535-9
Homeridai 492, 508, 541-82
Homicide 90, no, 132-7, 341
Homogalaktes 1 12
Horai 339-41, 484
Horde, primitive 39, 44, 61, 66, 69,
71
Howitt 50, 84-5
Hunting 21, 33-4, 42-4, 75, 88,
149-50, 204, 214, 297, 331
Hybrias 367
Hyksos 298, 412
Hypachaioi 386
Hypsipyle 122, 175, 277
Hyrnatheis 166
Hyrnetho 166, 560
Hysteria 227, 460
Iamidai 332-3
Ibykos 474
Idiean Cave 115
Ikarios 422, 426-8
Ilissos 248
Imbros 172-3
Immarados 128
Improvisation 436, 448, 454-62
Inachos 377
Incantations 245, 247, 441-3, 446,
467-8
Incest- 1 32, 417-8
India 36, 41, 60, 69, 113, 151-4,
183, 212, 303-5. 3H» 360, 431
GENERAL INDEX -
615
Indo-European: diaspora 176-7;
‘kin and ‘know’ 46, 93, 337;
'lot' 327; nomadism 298; 'path 1
134; ‘sea’ 172; terminology of
kinship 78-84, in, 145.
Infanticide 336
Initiation 45-9; African 207-8,
241-3; by bathing 225; by fire
504; paleolithic 56; Spartan 272.
Ino 263
Inspiration 437, 454-62
Investiture 482
Io 284-5, 377-82
Ion 126, 391-2
Ionian Is. 1x3, 312
Ionians 102-4, 390-2, 544-5;
early kings 165; women's costume
169.
Ionic dialect 197, 391, 515-26
Ioxidai 122, 164
Iphigeneia 338, 510
Iphitos 287
Iran 21, 25
Ireland 306, 540
Irish poetry 436, 443, 496, 566
Iroquois 87-92, 97, 100, 139, 150,
154. 214.
Ishtar 159, 180, 514
Isis 1 3 1, 159, 221, 284, 379
Isthmian Games 263
Italo-Celtic 80, 82
Ithaca 1 1 3, 420-6
Itonos 259, 397
Ivory Coast 1 57, 242
Japan 240
Jason 166, 175, 191, 195-6, 369
Jats 141
Javan 521
Jebb 41 5
Jews 138
Jolowicz 96
Joshua 327
Jukuns 50, 245
Juno 219, 290-1
Junod 50, 120, 303, 439, 447
Jupiter 290
* 337
Kadesh 401
Kadmeioi: in Ionian migration 391;
Phoenician origin 376-7.
Kadrnos, ancestor of Spartoi 121;
of Gephyraioi 123; in Elysium
513; Phoenician origin 376-7. See
Thebes, Boeotian.
Kaineidai 201
Kalauria 265
Kalchas 499
Kallinos 556, 560
Kambyses 162
Kanes 26
Kar 170, 377
Kardamyle 170, 395
Karpo 221
Karsten 503
Kassites 28, 298
Katharmata 207, 222-6, 263
Kaukones 171, 278
Keats 443-5
Kekrops 142-3, 176, 222, 257-8,
261-3, 266, 362-4
Keleai 129
Keos 350
Kephallenia 313, 421-3, 426-7
Kerameikos 364
Kerkyra 359
Kerykes 127-8, 172, 333
Khasis 15 1-4, 160, 272
Kilikes 401
Killos 386
Kimon 326
Kingship 23-4, 96, 156-8, 299,
328, 331, 346, 360, 364
Kinyras, Kinyradai 165, 513, 561
Kirghiz 459, 529-31, 54 °» 576
Kissioi 260-1
Kissousa 224
Klazomenai 129, 544
Kleisthenes of Athens 365
Kleisthenes of Sikyon 566
GENERAL INDEX
6l6
Kleomenes 117
Kliros 327, 333
Klerouchla 314, 323-5
Kleuas 543-4
Klytidai of Chios 125
Klytidai of Olympia 191
Knossos 29; fall of, 372-3, 383
Kodridai 186, 192, 198, 3 3 1 ,
544-6
Komana 407
KSme 351
Korakou 374
Kps 128, 167, 196, 202, 330-1
Kouretes 467
Kourotriphos 270
Kreousa 116
Kretheus 190-2
Kretschmer 163, 172, 260, 391
Kroeber 86
Krokonidai 127, 132
Kronos 325, 333-4
Kukis 349
Kupapa 290, 512
Kurgans 78
Kybebe 290, 5x2
Kybele 406-7
Kychreus 117, 389
Kydrolaos 326
Kyklopes 359
Kylon 133
Kynaithos 550, 554, 567, 570
Kynouria 393
Kypselos 201-2, 264-5, 565
Kyrene 101, 319-20, 323, 328, 561
Kythera 510
Laban 425
Labour service 297
Labour songs 446-8
Lachesis 325, 330, 334-5
Uclos 327, 333* 335
Laertes 422
Lakiadai 128, 326
Langdon 161
Lapithai 196, 197, 201, 353, 375 - 7 .
398-400, 523
Larisa 172, 257, 377. 3 82 » 542
Latin 431
Laughterless Rock 1 3 x
Lavinium 115
Lebedos 351, 544
Leda 428-9
Leleges 166-71, 425-30; Early
Cycladic and Helladic culture 177,
193, 374; of Lokris 199, 406; of
N.W. Anatolia 279, 406, 542; of
Sparta 271, 429.
Lemnos 172-5, 277-8, 571
Leontinoi 3x9
Lesbos: ^©olic settlement of 543-4;
Athenian plantation of 314-5;
Oinomaos king of 405; prehistoric
settlement of 326.
Lethaby 182
Leto 163, 229, 265, 293-4, 429
Letrinoi 223-4, 2 73
Leukadios 425
Leukas 360, 425
Leukippos 224
Leuktra 121, 257, 393, 395
Levirate 71
Libya 141, 232, 375, 379, 382, 40I
Limerick 496
Lion Gate 373, 407, 41 1
Liparai Is. 320-2
Livy 321
Lokris 170, 199, 259, 278, 406,
542-3
Lokroi 199, 227
Lorimer 381, 509
Lot 305, 314-5, 319-34, 347, 358
Lowie 36, 70, 86
Lucian 547
Lunda 156
Lyall 152
Lycia 163-6; crest 290, 371; Luka
401; matriarchal 98, 142, 144;
Proitos in 382.
Lydia 141, 174, 177-9, 2 7 i» 402,
486
Lykourgos 560
Magi 162, 271
Magnes 375
GENERAL INDEX
617
Maiden springs 245
Maiden’s Well 233
Maine 305
Makareus 325 —
Maiones 542
Malaos 543-4
Malaya 227
Malinowski 86, 441
Malis 353
Mandans 89
Manlii 94
Manrineia 351, 358-9
Maoris 185, 21 x, 440-1, 448
Marathon 264, 353
Market-place 354^ 362
Marlowe 301
Marr 176
Marx 85, 144, 151, 300-I
Masai 120
Massagetai 141, 298
Materialism 142, 348, 504, 529
Mattes Deae 343-4
Matricide 342
Mausolos 167
McLennan 144
Medontidai 364-5, 544
Megalopolis 234, 288, 251
Megaron 23 1
Melampous 226, 382
Melanesia 69, 72, 77, 119, 136, 2x3
Meleagros 342
Memphis 379, 508
Men 212
Menander 337
Menelaos 507-14; ancestry 408;
domain 393, 430.
Meriones 502
Mermnadai 174, 406
Mesopotamia 23-6
Matron 345
Mexico 92, 100, 338, 353
Meyer 259, 392, 517
Michell 308
Miletos 169, 294, 542-5
Millar 144, 300
Milon 219
Miltiades 572
Milton 527
Minnitaree 76
Minoa 177
Minos 369-70, 376', 427
Minotaur 285, 383
Minyas, Minyai 183-98, 263, 375,
390, 396, 399
Minyan ware 193, 375
Minyeios 191
Mohawks 236
Moiety 58-9, 69-73, 352
Moira 327-46
Molossoi 389
Moon-worship 210-18, 228-31,
279, 3 34 * 444
Morgan 43, 56, 59-66, 84-90, 100,
109, 144, 149
Mounychia 279
Mpongwe 213
Mttnius 232
Munro 266
Murray Is. 21 1
Muses 462, 484-5, 492-4, 496
Mycense 29, 249, 256, 280, 371-
84
Mylitta 294, 514
Myrmidons 387-8
Myrtilos 403-4, 406
Mysoi 542
Mytilene 18 1, 544-5
Names 46-7, 89, 92-4, 145
Nasamones 141
Naukratis 379, 381, 508, 545
Naxos 170, 202, 254
Nayars 183
Neleidai 192, 391-2, 398-9, 429,
545
Neoplatonism 228, 230
Nestor 121, 191, 361-2, 501
New Britain 43
New Hebrides 440, 442
Niebuhr 96
Nile 22, 150, 164
Nilsson 1 14-7, 197. 243 - 5 * 289,
293-4, 299, 300, 370, 381, 509-
10, 517-18
GENERAL INDEX
6l8
Niobe 402-6
NSmos 346-7
Norns 343-4
North-West Greek 399-400, 409
Novel 451-2
Nu-kuo 156
Obscenity in Greek ritual 206
Odysseus 420-5, 561
Oedipus 330, 342, 345, 560
Oths 109-11, 139, 153, 313-4, 323
Oinoe 353
Oinomaos 403-5
Olen 483
Olympias 217
Olympic Games 191, 269, 271, 282,
288, 292, 369, 403-4
Olympus 111, 362
Omphale 174, 290
Onomakritos 571
Orchomenos 187-98, 371-5
Orestes: ancestry 408; death 542.
Orgciti 112
Oropos 124, 398
Orphism 228-30, 337, 341, 504
Orthagoras 565
Ortygia 294
Orwin 307
Osage 76
Otreus 407
Oxylos 409
Paget 446
Palace cults 124-5, 193, 235, 255,
283
Palestine 26, 305, 318
Pamphos 483
Pandora 250
Pandrosos 261
Panionic League 390, 544
Panwar-Rajputs 43 1
Panyasis 167, 556
Paphos 413, 513
Parc® 343-4
Parens 98
Parian Marble 369, 371
Paris 383, 416, 507-9
Parry 534
Partheniai 200
Parthenogenesis 243, 284, 287
Pasiphae 255, 285
Pastoralism 33, 44, 51, 7 8, 149,
298, 349
Patesi 161
Path-finding 134
Pavlov 52
Pearson 558
Pedasos 170
Peirithoos, Peirithoidai 264-5
Peisistratos, Peisistratidai 121, 192,
264, 548, 568-75
Pelasgoi 171-7, 257-61, 277-9 ; in
Ionian migration 391; Minyan
ware 193, 375; orgeSn 1 13.
Peleus 387-9, 395-7
Pelops 402-1 1
Pentheus 486
Penthilos, Penthilidai 408-10, 542,
546.
Periandros 265, 565
Persephone 231-7: see Demeter.
Perseus 369, 379-84, 427-8
Persia 25, 169
Petra 197, 201, 264-5, 400
Petrie 160
Phaeacia 359-62, 418, 430
Pharos 507-9 -
Pheidias 268
Pheidon 561
Phemios 459, 494
Pheneos 131, 257
Pheros 508, 512
Philaios, Philaidai 121, 169, 264,
326, 572
Philippine Is. 240
Philistines 164, 172
Philochoros 112
Phleious 129, 544
Phoenicians 119, 124, 320-2, 376-
9, 384, 427, 502, 508-12
Phoinix 414-5
Phokos 388-9, 395, 400
Phorbas 375
GENERAL INDEX
Phoroneus 377
Phratry 58-9,71-3, S7-90, X04-S,
145 . 3 * 3 - 3 2 5 . 352
Phtygia 232, 235, 406-7
Phylakc 192, 398
Picard 269-71, 294
Pindar 465, 474, 480-2
Piracy 322, 329, 360
Pisa 224, 404-5
Pitanc 1 8 1 , 429, 542
Plato: on Dcmokritos 143; ideal city
324; on Homer 458, 550, 557;
on inspiration 459.
Plautus 141
PIcistlicncs 409
Plcuron 406, 427-S
Plough 42, 307, 309, 317
Plough-oxen 51, 122
PeUmatths 364
PJlit 348-52, 558, 364
Polyandry 71, 143
Polynesia 60-1, 64-6, 69
Pomegranate 219, 237, 252, 288
Pempa{ 122, 196
Porphyry 230
Poseidon: contest with Athena 262-
3, 266; father of Boutes 126, 265-
6; of Eumolpos 127; of Taphios
427; god of Lapithai 264-5;
Hclikonios 104, 390; Petraios
197; at Pylos 362; sea-god ill.
Possession 459-60
Potniai 129
Pottery' 21, 33-4
Praxiergidai 482
Praxilla 498
Pre-nuptial contest 49
Pre-nuptial promiscuity 141
Priam 259-61, 328, 387, 389, 407,
416-7
Procrosia 309
Profession of vows 46-7
Proitos 191, 226, 284, 382-3
Proklos 552
Prometheus 370
Prophecy 215, 460, 485
Protesilaos 397
619
Proteus 507-9
Prytancton 363
Ptolemies 162
Purification 46, 48, 207, 226, 482
Pyrgo: 169
Pythagoreans 337
Py thermos 496
Quince 288
Rachmani 190
Radcliffe-Brown 85
Radtov 459, 530-1, 540, 576
Rain-making 1 57-8
Rebirth 45-7, 95, 109, 229, 333,
404, 419
Rcichclt 520
Revillouc 160
Rhadamanthys 341, 359, 513
Rhapsode 48S, 491, 574-5, 577
Rhea 235, 245, 256
Rhodes: Achaans in 386; dialect
399, 517, 524; Kadmos and
Danaos 376, 379; tribal settlement
324-6, 353; urbanisation 350-1.
Rhyme 465, 470
Riddle 499
Rivers 64
Robertson Smith 36, 161, 212, 336,
356
Romans 25, 92-101, 139, 154, 205,
232, 290-1, 431
Romulus 94, 97, 101
Roschcr 228, 237
Rostovtzeff 144
Russian poetry 448, 457
Sadazios 235
Sabines 99
Sais 379
Sacrament 51
Salamis u8, 169, 386, 389, 572
Samos: Carians in 170; dialect 544;
feast of Adonis at 499; name 427;
settlement of 326; tribal system
106; temple of Hera 218, 282-3.
GENERAL INDEX
620
Samothracian Mysteries 173
Sandas 290, 512
Santa Mavra 313
Sappho 248, 452-3, 465-74, 498
Sargon 164
Sarpedon 164, 177, 346
Schefold 365
Scheria: see Phasacia
Schliemann 193, 239, 244, 529
Scotland 306
Scott 525-6, 528
Scythians 140
Seasons 236, 339
Second wife 160
Secret societies 51, 89, 213
Seebohm 111, 306
Seleucids 162
Self-help, primitive 1 32-3
Selloi 397-8
Semele 339
Semiramis 162, 183
Semites 36, 119, 138, 161, 213
Semitic 60, 204, 212
Senecas 88, 90
Sennacherib 521
Serfdom 298, 422
Servius 131
Shaft Graves 371, 374, 383, 502
Shakespeare 453, 457, 580
Shelley 577
Siberia 240
Sicily 29, 126-7, 209, 233, 359,
374
Sigeion 569
Sikyon 235, 280, 498, 565-6
Simonides 474
Sinai 25
Sinclair 300, 322
Sipylos 402, 406
Sirens 285
Sisyphos 164, 190-1
Skimismal 288
Skirophoria 221-2
Slavery. 89, 142, 355, 358
Smith, Adam 144, 300
Smyrna 181, 257, 515, 525-6, 542,
545
Solon 312, 319, 474, 481, 569
Sophokles 173-4, 383
Sororate 71
Sosipolis 1 15
Sousa 260, 277
Soviet Union 41, 85-6, 144, 301
Spain 320, 374
Spartoi 121, 3"37
Spencer and Gillen 37, 75, 84
Springs 224, 245
Stddion 3x8
Stanza 441, 464-74
Staweill 525-6
Stesichoros 465-6, 474, 508, 567
Stoicism 21 5
Strabo, on formation of towns 350.
Suidas 547
Sumer 161, 356
Sumptuary laws 481
Svayamvara 404, 580
Symphony 452
Syracuse 126-7, 466, 567
Syria 376, 485, 512-13
Taboo 36-40, 44, 51, 119, 122,
152, 204-5, 209, 220, 222, 250
Tacitus 308, 527
Tanistry 97
Tantalos 402-3
Taphioi 423, 426-7
Taras 200-1
Tarchon 174, 179
Tarn 162
Tarquinius 97, 174, 179
Tattoo 337
Tauros 26, 180
Tegea 132, 201
Telamon 166, 387-8
Teleboas, Teleboai 426-7.
Telephos 174
Temenos 166 -
Timenos 326, 329-31, 346, 355-7,.
360, 365, 404
Tenedos 278, 326, 542
Teos 169, 326, 351, 496, 544 -
Ternary form 451-3
Terpandros 466, 492
GENERAL INDEX
621
Teshub 1 80
Testamentary disposition 93, no,
312
Teukroi 386-7
Teukros, Teukridai 386-7, 561
Tcutamos 260, 406
Tlalassa 172, 176, 263
Thaleia 3 39
Theano 260
Thebe 397
Thebes: Boeotian, foundation 124,
376; subject to Orchomenos 187.
See Kadmos, Kadmeioi.
Thebes, Egyptian 380, 509-10
Thcmistokles 330
Theopompos 142
Therapnc 170, 513
Theseus 122, 254, 264-6, 326,
362-5, 369, 573
Thesmophoria 124, 220-3, 231-2,
236, 273
Thespiai 129, 485
Thessaloi 259, 396, 524
Thetis 389, 403, 484, 492
Thisbe 374
Tholos Tomb Dynasty 371, 380-3
Tholos tombs 249
Thomsen 176
Thrace 127, 172, 260
Thucydides: on Carians 170; early
Greece 142, 257, 348-9, 527;
Lesbos 315; Pelasgoi 173; Sparta
316; Theseus 362-3; Trojan War
348, 515.
Tiberius 216
Tibet 156
Tikopia 61, 441
Timi 329, 333
Tityns 29, 164, 282, 371, 374, 377*
382, 404
Tisamenos 408-9
Titans 220, 333
Tithes 357
Tithonos 260-1
Tlepolemos 324-5
Tokens 336-7
Tolstoy 295
Tombs 371-4, 381, 407, 410
Tools 437-8, 451
Tooth evulsion 47-8, 56, 213
Totemism 36-57, 66, 86-9, 94,
114-23, 250, 276, 289, 337-8,
500
Toutain 301-2, 320-1
Town-hall 363
Treasury of Atreus 374, 41 1
Triakds 106
Triopas 375
Triptolemos 128-9, 131, 256, 325
Trobriand Is. 441
Troy: fall of 369; site 29, 541.
Turkmens 458
Tyro, Tyroidai 191-2, 198, 375,
382, 400
Tyrol 242
Tyrrhcnos 174
Ugarit 26, 376-7
Ukraine 78
Usener 474
Uterine posture: see Burial,
contracted
Valenge 209, 218, 241-3, 250
Varro 232
Vendetta no, 481
Vendrycs 80
Venus: of Laussel 344; of Willen-
dorf 237.
Vestals 158
Vico 528
Village community zi, 248, 304-8,
314 . 353 - 5 . 357-8
Vinogradoff 302
Virgil 527
Voluspd 367
Wace 371
Wackemagel 526
Wade-Gery 202, 554, 570
622 GENERAL INDEX
Wales 136, 306
Warfare 298, 355, 500
Webster 45
Westermarck 70
Wife-lending 143
Wilamowitz 334, 515, 528
Winnebagos 90
Witchcraft 132, 213, 229
Wolf 528, 574
Wooden Horse 494
Woolley 376
Xenophon 557
Xerxes 162, 169
Xouthos 391
Yeats 433, 455
Ydkumbil 43
Zakynthos 353, 385, 421, 423
Zante 313
Zeus: suckled by Amaltheia 256;
Atabyrios 178; father of Athena
267; birth 245; marriage to Dione
292; at Dodona 171, 292, 397,
400; rape of Europa 123, 376;
Geleon 104; goat 250; marriage to
Hera 280-6; Karios 170; Herkeios
313;- Labrandeus 290; Leukaios
224; name 250, 400; Olympios
413-4; Patroios 125; Pelasgios
171, 397; Saviour 495; sky-god
hi; sow 250; swan 429, 506.
Zulus 119, 134, 206, 208, 439 '
Zygioi 387
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11